CIFAR_CNN_Practice.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CIFAR-CNN-Practice.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyM6LSDz5cLMQFd07rBU4nms",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/aayushkubb/Deep_Learning_Tutorial/blob/master/CIFAR_CNN_Practice.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RTlfacWY_nVv"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import tensorflow as tf\n",
        "sns.set()\n"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N2-WSmhtCcak"
      },
      "source": [
        "from keras.datasets import cifar10\n",
        "import keras\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gg6qGWBlCeEx",
        "outputId": "c4f214f7-9a40-464d-f36a-4aec290774c7"
      },
      "source": [
        "(x_train,y_train),(x_test,y_test)=cifar10.load_data()"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n",
            "170500096/170498071 [==============================] - 2s 0us/step\n",
            "170508288/170498071 [==============================] - 2s 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0Hmn7ZVBCtgp",
        "outputId": "eb8c9e49-a967-45d9-9178-4df9f4ede57b"
      },
      "source": [
        "x_train.shape"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(50000, 32, 32, 3)"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Bkb_dZIcEYAf",
        "outputId": "d1c456d8-ee83-45b4-bcea-0ccf552013ba"
      },
      "source": [
        "y_train.shape"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(50000, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GRgbkzHLEbdq",
        "outputId": "b4a95779-14f9-4645-cb93-7489090197b9"
      },
      "source": [
        "np.unique(y_train)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=uint8)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 262
        },
        "id": "W7iDaCsbEds8",
        "outputId": "92ec08f8-97aa-45a5-cb5c-34bd23dd2048"
      },
      "source": [
        "print('Training data shape : ', x_train.shape, y_train.shape)\n",
        "\n",
        "print('Testing data shape : ', x_test.shape, y_test.shape)\n",
        "\n",
        "# Find the unique numbers from the train labels\n",
        "classes = np.unique(y_train)\n",
        "nClasses = len(classes)\n",
        "print('Total number of outputs : ', nClasses)\n",
        "print('Output classes : ', classes)\n",
        "\n",
        "plt.figure(figsize=[4,2])\n",
        "\n",
        "# Display the first image in training data\n",
        "plt.subplot(121)\n",
        "plt.imshow(x_train[0,:,:], cmap='gray')\n",
        "plt.title(\"Ground Truth : {}\".format(y_train[0]))\n",
        "\n",
        "# Display the first image in testing data\n",
        "plt.subplot(122)\n",
        "plt.imshow(x_test[0,:,:], cmap='gray')\n",
        "plt.title(\"Ground Truth : {}\".format(y_test[0]))"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training data shape :  (50000, 32, 32, 3) (50000, 1)\n",
            "Testing data shape :  (10000, 32, 32, 3) (10000, 1)\n",
            "Total number of outputs :  10\n",
            "Output classes :  [0 1 2 3 4 5 6 7 8 9]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Ground Truth : [3]')"
            ]
          },
          "metadata": {},
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 288x144 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fKy0XBrEE1oy"
      },
      "source": [
        "# Prepare data for training\n",
        "nRows,nCols,nDims=x_train.shape[1:]"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Upnxl2x5FLGn"
      },
      "source": [
        "# Make all images of same shape and size\n",
        "x_train=x_train.reshape(x_train.shape[0],nRows,nCols,nDims)\n",
        "x_test=x_test.reshape(x_test.shape[0],nRows,nCols,nDims)"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3mVgPfHEFqmd",
        "outputId": "dbb1614c-a124-405d-da04-6d83f5c37034"
      },
      "source": [
        "x_train.shape, x_test.shape"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((50000, 32, 32, 3), (10000, 32, 32, 3))"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L4SXYLM1FtJ7"
      },
      "source": [
        "# Scale the data\n",
        "x_train=x_train.astype('float32')\n",
        "x_test=x_test.astype('float32')"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "G5F79mhdF9B-",
        "outputId": "e5e7a260-6c5f-47f1-d378-4fc52d36b1da"
      },
      "source": [
        "np.max(x_train[0]),np.min(x_train[0])"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(255.0, 0.0)"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7iV8A5l2GBGj"
      },
      "source": [
        "# Scale the data\n",
        "x_train=x_train/255\n",
        "x_test=x_test/255"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GXjmGk3bGSKT",
        "outputId": "7b7df5cc-0918-4159-c0fb-0ca736c40469"
      },
      "source": [
        "np.max(x_train[0]),np.min(x_train[0])"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1.0, 0.0)"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4lyXMOtcGTMW"
      },
      "source": [
        "from tensorflow.keras.utils import to_categorical"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L6f5jkt1Gs9X"
      },
      "source": [
        "y_train_labels=to_categorical(y_train)\n",
        "y_test_labels=to_categorical(y_test)"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GhqCxww6G0lc",
        "outputId": "5a04ba0b-238b-4ec6-cc66-74d04af19c30"
      },
      "source": [
        "y_train_labels.shape"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(50000, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7vBU8owXG1xB",
        "outputId": "0b2ecde5-5faf-412a-a45d-797a5e37377d"
      },
      "source": [
        "y_train_labels[0]"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0., 0., 0., 0., 0., 0., 1., 0., 0., 0.], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dcvE3h5SG39e",
        "outputId": "13447bdd-6b80-4b3f-ec87-46a6c38f63a9"
      },
      "source": [
        "y_train[0]"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([6], dtype=uint8)"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lOHDLZNRHdKN"
      },
      "source": [
        "from keras.layers import Conv2D,Flatten"
      ],
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZcgefuZ3H00p"
      },
      "source": [
        "def __init__(filters, kernel_size, strides=(1, 1), padding='valid', data_format=None, dilation_rate=(1, 1), groups=1, activation=None, use_bias=True, kernel_initializer='glorot_uniform', bias_initializer='zeros', kernel_regularizer=None, bias_regularizer=None, activity_regularizer=None, kernel_constraint=None, bias_constraint=None, **kwargs)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zwFvvdfSIPb9",
        "outputId": "91dacf61-95f7-44b5-baaf-3db0b2b7b436"
      },
      "source": [
        "x_train[0].shape"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(32, 32, 3)"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RCOfulVUG6Kr"
      },
      "source": [
        "model=Sequential()\n",
        "model.add(Conv2D(32,(3,3),padding='same',activation='relu',input_shape=(32, 32, 3)))\n",
        "model.add(Conv2D(32,(3,3),padding='valid',activation='relu'))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(512,activation='relu'))\n",
        "model.add(Dense(10,activation='softmax'))"
      ],
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BIWNWKVfIS0B",
        "outputId": "fbf61542-ad56-4b96-d06b-6d4f970536bc"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 32, 32, 32)        896       \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 30, 30, 32)        9248      \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 28800)             0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 512)               14746112  \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 10)                5130      \n",
            "=================================================================\n",
            "Total params: 14,761,386\n",
            "Trainable params: 14,761,386\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NlolsNYUJQWA"
      },
      "source": [
        "model.compile(optimizer='rmsprop',loss='categorical_crossentropy',metrics=['accuracy'])"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8NLu5jVdIx98",
        "outputId": "a0402749-cabf-451e-bc87-08e44f19c05b"
      },
      "source": [
        "history=model.fit(x_train,y_train_labels,epochs=5,batch_size=512,verbose=1,validation_data=(x_test,y_test_labels))"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "98/98 [==============================] - 35s 75ms/step - loss: 2.1415 - accuracy: 0.3510 - val_loss: 1.5154 - val_accuracy: 0.4569\n",
            "Epoch 2/5\n",
            "98/98 [==============================] - 7s 68ms/step - loss: 1.4394 - accuracy: 0.4992 - val_loss: 1.2800 - val_accuracy: 0.5380\n",
            "Epoch 3/5\n",
            "98/98 [==============================] - 7s 68ms/step - loss: 1.2085 - accuracy: 0.5783 - val_loss: 1.1736 - val_accuracy: 0.5898\n",
            "Epoch 4/5\n",
            "98/98 [==============================] - 7s 68ms/step - loss: 1.0231 - accuracy: 0.6449 - val_loss: 1.2443 - val_accuracy: 0.5641\n",
            "Epoch 5/5\n",
            "98/98 [==============================] - 7s 69ms/step - loss: 0.8733 - accuracy: 0.6980 - val_loss: 1.5757 - val_accuracy: 0.4983\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "c3qKA67iJN9y",
        "outputId": "14a9d7d3-71ab-4af6-d193-ee49764b8c25"
      },
      "source": [
        "model.evaluate(x_test,y_test_labels)"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 2s 5ms/step - loss: 1.5757 - accuracy: 0.4983\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.5757131576538086, 0.4982999861240387]"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dKnv1sJPKOKj"
      },
      "source": [
        "from keras.preprocessing.image import ImageDataGenerator"
      ],
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KH4M2epGJyOW"
      },
      "source": [
        "model=Sequential()\n",
        "model.add(Conv2D(32,(3,3),padding='same',activation='relu',input_shape=(32, 32, 3)))\n",
        "model.add(Conv2D(32,(3,3),padding='valid',activation='relu'))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(512,activation='relu'))\n",
        "model.add(Dense(10,activation='softmax'))\n",
        "model.compile(optimizer='rmsprop',loss='categorical_crossentropy',metrics=['accuracy'])\n"
      ],
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uMJl1gIAKkCY"
      },
      "source": [
        "datagen=ImageDataGenerator(horizontal_flip=True,vertical_flip=True)"
      ],
      "execution_count": 29,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BhWh9regLDnT",
        "outputId": "50e0ce32-fa41-494a-b52b-ea9cf7d56e26"
      },
      "source": [
        "hostory2=model.fit_generator(datagen.flow(x_train,y_train_labels),steps_per_epoch=10,epochs=100,validation_data=(x_test,y_test_labels))"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            " 7/10 [====================>.........] - ETA: 0s - loss: 1.3249 - accuracy: 0.5357"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/engine/training.py:1972: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
            "  warnings.warn('`Model.fit_generator` is deprecated and '\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3051 - accuracy: 0.5437 - val_loss: 1.5371 - val_accuracy: 0.4677\n",
            "Epoch 2/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3762 - accuracy: 0.5656 - val_loss: 1.3628 - val_accuracy: 0.5106\n",
            "Epoch 3/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3540 - accuracy: 0.4781 - val_loss: 1.3615 - val_accuracy: 0.5126\n",
            "Epoch 4/100\n",
            "10/10 [==============================] - 2s 167ms/step - loss: 1.3888 - accuracy: 0.5125 - val_loss: 1.3440 - val_accuracy: 0.5188\n",
            "Epoch 5/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.4172 - accuracy: 0.5281 - val_loss: 1.3585 - val_accuracy: 0.5066\n",
            "Epoch 6/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3221 - accuracy: 0.5125 - val_loss: 1.4639 - val_accuracy: 0.4780\n",
            "Epoch 7/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.3806 - accuracy: 0.5000 - val_loss: 1.3564 - val_accuracy: 0.5163\n",
            "Epoch 8/100\n",
            "10/10 [==============================] - 2s 167ms/step - loss: 1.2718 - accuracy: 0.5250 - val_loss: 1.5005 - val_accuracy: 0.4732\n",
            "Epoch 9/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3907 - accuracy: 0.5031 - val_loss: 1.3533 - val_accuracy: 0.5172\n",
            "Epoch 10/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.4006 - accuracy: 0.4906 - val_loss: 1.3595 - val_accuracy: 0.5153\n",
            "Epoch 11/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.2384 - accuracy: 0.5625 - val_loss: 1.3581 - val_accuracy: 0.5186\n",
            "Epoch 12/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3565 - accuracy: 0.5469 - val_loss: 1.4146 - val_accuracy: 0.4982\n",
            "Epoch 13/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.4098 - accuracy: 0.5125 - val_loss: 1.3419 - val_accuracy: 0.5213\n",
            "Epoch 14/100\n",
            "10/10 [==============================] - 3s 303ms/step - loss: 1.2516 - accuracy: 0.5562 - val_loss: 1.4048 - val_accuracy: 0.5059\n",
            "Epoch 15/100\n",
            "10/10 [==============================] - 2s 167ms/step - loss: 1.3313 - accuracy: 0.5281 - val_loss: 1.3485 - val_accuracy: 0.5183\n",
            "Epoch 16/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3875 - accuracy: 0.5031 - val_loss: 1.3116 - val_accuracy: 0.5376\n",
            "Epoch 17/100\n",
            "10/10 [==============================] - 3s 303ms/step - loss: 1.3373 - accuracy: 0.5219 - val_loss: 1.4222 - val_accuracy: 0.4892\n",
            "Epoch 18/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.3065 - accuracy: 0.5437 - val_loss: 1.4083 - val_accuracy: 0.5055\n",
            "Epoch 19/100\n",
            "10/10 [==============================] - 2s 172ms/step - loss: 1.2642 - accuracy: 0.5625 - val_loss: 1.4398 - val_accuracy: 0.4968\n",
            "Epoch 20/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.2512 - accuracy: 0.5688 - val_loss: 1.4392 - val_accuracy: 0.4983\n",
            "Epoch 21/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3277 - accuracy: 0.5031 - val_loss: 1.4182 - val_accuracy: 0.4996\n",
            "Epoch 22/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3795 - accuracy: 0.5250 - val_loss: 1.3273 - val_accuracy: 0.5316\n",
            "Epoch 23/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.2467 - accuracy: 0.5625 - val_loss: 1.4509 - val_accuracy: 0.4918\n",
            "Epoch 24/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.3447 - accuracy: 0.5188 - val_loss: 1.4214 - val_accuracy: 0.4987\n",
            "Epoch 25/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.3499 - accuracy: 0.5469 - val_loss: 1.4882 - val_accuracy: 0.4850\n",
            "Epoch 26/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3577 - accuracy: 0.5188 - val_loss: 1.3085 - val_accuracy: 0.5282\n",
            "Epoch 27/100\n",
            "10/10 [==============================] - 2s 169ms/step - loss: 1.2650 - accuracy: 0.5281 - val_loss: 1.5190 - val_accuracy: 0.4851\n",
            "Epoch 28/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.2980 - accuracy: 0.5625 - val_loss: 1.3321 - val_accuracy: 0.5316\n",
            "Epoch 29/100\n",
            "10/10 [==============================] - 3s 303ms/step - loss: 1.3778 - accuracy: 0.5781 - val_loss: 1.3890 - val_accuracy: 0.5047\n",
            "Epoch 30/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.3297 - accuracy: 0.5437 - val_loss: 1.3966 - val_accuracy: 0.4967\n",
            "Epoch 31/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3450 - accuracy: 0.5406 - val_loss: 1.3192 - val_accuracy: 0.5287\n",
            "Epoch 32/100\n",
            "10/10 [==============================] - 2s 167ms/step - loss: 1.2861 - accuracy: 0.5437 - val_loss: 1.2972 - val_accuracy: 0.5391\n",
            "Epoch 33/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.1963 - accuracy: 0.6031 - val_loss: 1.3028 - val_accuracy: 0.5388\n",
            "Epoch 34/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3477 - accuracy: 0.5469 - val_loss: 1.3067 - val_accuracy: 0.5374\n",
            "Epoch 35/100\n",
            "10/10 [==============================] - 3s 303ms/step - loss: 1.2801 - accuracy: 0.5406 - val_loss: 1.2884 - val_accuracy: 0.5416\n",
            "Epoch 36/100\n",
            "10/10 [==============================] - 2s 172ms/step - loss: 1.2809 - accuracy: 0.5250 - val_loss: 1.5747 - val_accuracy: 0.4586\n",
            "Epoch 37/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.3339 - accuracy: 0.5312 - val_loss: 1.3554 - val_accuracy: 0.5178\n",
            "Epoch 38/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3906 - accuracy: 0.4969 - val_loss: 1.3933 - val_accuracy: 0.5173\n",
            "Epoch 39/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.2953 - accuracy: 0.5375 - val_loss: 1.3065 - val_accuracy: 0.5369\n",
            "Epoch 40/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.2601 - accuracy: 0.5437 - val_loss: 1.3708 - val_accuracy: 0.5098\n",
            "Epoch 41/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.2542 - accuracy: 0.5469 - val_loss: 1.3027 - val_accuracy: 0.5349\n",
            "Epoch 42/100\n",
            "10/10 [==============================] - 1s 163ms/step - loss: 1.3325 - accuracy: 0.5469 - val_loss: 1.3174 - val_accuracy: 0.5301\n",
            "Epoch 43/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3292 - accuracy: 0.5469 - val_loss: 1.3706 - val_accuracy: 0.5166\n",
            "Epoch 44/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3411 - accuracy: 0.5437 - val_loss: 1.4199 - val_accuracy: 0.5006\n",
            "Epoch 45/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.3238 - accuracy: 0.5312 - val_loss: 1.3460 - val_accuracy: 0.5118\n",
            "Epoch 46/100\n",
            "10/10 [==============================] - 1s 163ms/step - loss: 1.3472 - accuracy: 0.5250 - val_loss: 1.3709 - val_accuracy: 0.5062\n",
            "Epoch 47/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.2457 - accuracy: 0.5437 - val_loss: 1.3560 - val_accuracy: 0.5164\n",
            "Epoch 48/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.3062 - accuracy: 0.5375 - val_loss: 1.3365 - val_accuracy: 0.5233\n",
            "Epoch 49/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.2727 - accuracy: 0.5656 - val_loss: 1.3775 - val_accuracy: 0.5094\n",
            "Epoch 50/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.2791 - accuracy: 0.5500 - val_loss: 1.3996 - val_accuracy: 0.5077\n",
            "Epoch 51/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.3108 - accuracy: 0.5500 - val_loss: 1.4273 - val_accuracy: 0.4972\n",
            "Epoch 52/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.2853 - accuracy: 0.4969 - val_loss: 1.3389 - val_accuracy: 0.5277\n",
            "Epoch 53/100\n",
            "10/10 [==============================] - 2s 171ms/step - loss: 1.3461 - accuracy: 0.5500 - val_loss: 1.3563 - val_accuracy: 0.5257\n",
            "Epoch 54/100\n",
            "10/10 [==============================] - 2s 167ms/step - loss: 1.2255 - accuracy: 0.5500 - val_loss: 1.2905 - val_accuracy: 0.5383\n",
            "Epoch 55/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3401 - accuracy: 0.5344 - val_loss: 1.2915 - val_accuracy: 0.5441\n",
            "Epoch 56/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3034 - accuracy: 0.5656 - val_loss: 1.3773 - val_accuracy: 0.5189\n",
            "Epoch 57/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.1898 - accuracy: 0.5781 - val_loss: 1.3362 - val_accuracy: 0.5215\n",
            "Epoch 58/100\n",
            "10/10 [==============================] - 3s 305ms/step - loss: 1.1982 - accuracy: 0.5969 - val_loss: 1.3858 - val_accuracy: 0.5148\n",
            "Epoch 59/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.2294 - accuracy: 0.5688 - val_loss: 1.4597 - val_accuracy: 0.4922\n",
            "Epoch 60/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.2407 - accuracy: 0.5875 - val_loss: 1.2742 - val_accuracy: 0.5455\n",
            "Epoch 61/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3555 - accuracy: 0.5500 - val_loss: 1.3669 - val_accuracy: 0.5176\n",
            "Epoch 62/100\n",
            "10/10 [==============================] - 1s 163ms/step - loss: 1.3029 - accuracy: 0.5531 - val_loss: 1.3690 - val_accuracy: 0.5170\n",
            "Epoch 63/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.2736 - accuracy: 0.5625 - val_loss: 1.3643 - val_accuracy: 0.5205\n",
            "Epoch 64/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.2351 - accuracy: 0.5813 - val_loss: 1.2615 - val_accuracy: 0.5479\n",
            "Epoch 65/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.2330 - accuracy: 0.5781 - val_loss: 1.3769 - val_accuracy: 0.5249\n",
            "Epoch 66/100\n",
            "10/10 [==============================] - 3s 305ms/step - loss: 1.2137 - accuracy: 0.5875 - val_loss: 1.3069 - val_accuracy: 0.5470\n",
            "Epoch 67/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.1818 - accuracy: 0.6031 - val_loss: 1.6149 - val_accuracy: 0.4552\n",
            "Epoch 68/100\n",
            "10/10 [==============================] - 2s 167ms/step - loss: 1.1975 - accuracy: 0.5906 - val_loss: 1.2877 - val_accuracy: 0.5388\n",
            "Epoch 69/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.3778 - accuracy: 0.4844 - val_loss: 1.2800 - val_accuracy: 0.5367\n",
            "Epoch 70/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.2790 - accuracy: 0.5781 - val_loss: 1.2495 - val_accuracy: 0.5553\n",
            "Epoch 71/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3001 - accuracy: 0.5969 - val_loss: 1.4113 - val_accuracy: 0.4988\n",
            "Epoch 72/100\n",
            "10/10 [==============================] - 1s 163ms/step - loss: 1.2575 - accuracy: 0.6062 - val_loss: 1.2969 - val_accuracy: 0.5402\n",
            "Epoch 73/100\n",
            "10/10 [==============================] - 1s 161ms/step - loss: 1.2176 - accuracy: 0.5688 - val_loss: 1.2506 - val_accuracy: 0.5507\n",
            "Epoch 74/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.3670 - accuracy: 0.4875 - val_loss: 1.3277 - val_accuracy: 0.5250\n",
            "Epoch 75/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.1855 - accuracy: 0.5906 - val_loss: 1.2307 - val_accuracy: 0.5647\n",
            "Epoch 76/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.2810 - accuracy: 0.5344 - val_loss: 1.2993 - val_accuracy: 0.5433\n",
            "Epoch 77/100\n",
            "10/10 [==============================] - 1s 162ms/step - loss: 1.2811 - accuracy: 0.5531 - val_loss: 1.2916 - val_accuracy: 0.5395\n",
            "Epoch 78/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.2245 - accuracy: 0.5969 - val_loss: 1.3795 - val_accuracy: 0.5216\n",
            "Epoch 79/100\n",
            "10/10 [==============================] - 2s 164ms/step - loss: 1.2840 - accuracy: 0.5437 - val_loss: 1.3273 - val_accuracy: 0.5251\n",
            "Epoch 80/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.2699 - accuracy: 0.5531 - val_loss: 1.3300 - val_accuracy: 0.5353\n",
            "Epoch 81/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.1831 - accuracy: 0.5969 - val_loss: 1.2676 - val_accuracy: 0.5482\n",
            "Epoch 82/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.3244 - accuracy: 0.5031 - val_loss: 1.3139 - val_accuracy: 0.5294\n",
            "Epoch 83/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.1872 - accuracy: 0.5719 - val_loss: 1.4658 - val_accuracy: 0.5078\n",
            "Epoch 84/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3023 - accuracy: 0.5500 - val_loss: 1.2812 - val_accuracy: 0.5498\n",
            "Epoch 85/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.1963 - accuracy: 0.5656 - val_loss: 1.3480 - val_accuracy: 0.5284\n",
            "Epoch 86/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.2410 - accuracy: 0.5844 - val_loss: 1.3210 - val_accuracy: 0.5357\n",
            "Epoch 87/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.2214 - accuracy: 0.5437 - val_loss: 1.2688 - val_accuracy: 0.5539\n",
            "Epoch 88/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.1137 - accuracy: 0.6094 - val_loss: 1.2964 - val_accuracy: 0.5421\n",
            "Epoch 89/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.2088 - accuracy: 0.5344 - val_loss: 1.3050 - val_accuracy: 0.5418\n",
            "Epoch 90/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.2286 - accuracy: 0.6062 - val_loss: 1.2648 - val_accuracy: 0.5454\n",
            "Epoch 91/100\n",
            "10/10 [==============================] - 2s 167ms/step - loss: 1.2608 - accuracy: 0.5531 - val_loss: 1.2934 - val_accuracy: 0.5431\n",
            "Epoch 92/100\n",
            "10/10 [==============================] - 2s 163ms/step - loss: 1.1685 - accuracy: 0.5813 - val_loss: 1.3106 - val_accuracy: 0.5403\n",
            "Epoch 93/100\n",
            "10/10 [==============================] - 1s 162ms/step - loss: 1.2581 - accuracy: 0.5562 - val_loss: 1.2583 - val_accuracy: 0.5518\n",
            "Epoch 94/100\n",
            "10/10 [==============================] - 2s 165ms/step - loss: 1.3070 - accuracy: 0.5531 - val_loss: 1.3170 - val_accuracy: 0.5400\n",
            "Epoch 95/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.1713 - accuracy: 0.5938 - val_loss: 1.4201 - val_accuracy: 0.5132\n",
            "Epoch 96/100\n",
            "10/10 [==============================] - 3s 304ms/step - loss: 1.2474 - accuracy: 0.5312 - val_loss: 1.2721 - val_accuracy: 0.5404\n",
            "Epoch 97/100\n",
            "10/10 [==============================] - 2s 168ms/step - loss: 1.2347 - accuracy: 0.5625 - val_loss: 1.3086 - val_accuracy: 0.5401\n",
            "Epoch 98/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.1527 - accuracy: 0.6375 - val_loss: 1.3843 - val_accuracy: 0.5114\n",
            "Epoch 99/100\n",
            "10/10 [==============================] - 2s 166ms/step - loss: 1.1761 - accuracy: 0.5750 - val_loss: 1.3699 - val_accuracy: 0.5217\n",
            "Epoch 100/100\n",
            "10/10 [==============================] - 3s 305ms/step - loss: 1.1032 - accuracy: 0.6031 - val_loss: 1.3029 - val_accuracy: 0.5540\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "32-WJ1ItMJsj"
      },
      "source": [
        "# Add maxpooling layers\n",
        "# Add dropouts\n",
        "# Create Dense layers\n",
        "\n",
        "# Experiments\n",
        "# Change filter size(big/small)\n",
        "# Change dense layer size\n",
        "# Batch Normalization\n",
        "# Batch Normalization with Dropouts\n",
        "\n",
        "\n",
        "# RESNET"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A5fKsPv8OfDQ"
      },
      "source": [
        "from keras.callbacks import ModelCheckpoint"
      ],
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BKFWL1Y9REWI"
      },
      "source": [
        "model.compile(optimizer='rmsprop',loss='categorical_crossentropy',metrics=['accuracy'])"
      ],
      "execution_count": 44,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "r0Pg4uteOgjU",
        "outputId": "6c7f6566-ee8d-4202-ea1c-10a2d53ae613"
      },
      "source": [
        "filepath = 'my_best_model.epoch{epoch:02d}-loss{val_loss:.2f}.hdf5'\n",
        "checkpoint = ModelCheckpoint(filepath=filepath, \n",
        "                             monitor='val_loss',\n",
        "                             verbose=1, \n",
        "                             save_best_only=True,\n",
        "                             mode='min')\n",
        "callbacks = [checkpoint]\n",
        "model.fit(x_train,\n",
        "          y_train_labels,\n",
        "          epochs=5,\n",
        "          callbacks=callbacks,validation_data=(x_test,y_test_labels))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "1563/1563 [==============================] - 31s 20ms/step - loss: 0.7523 - accuracy: 0.7478 - val_loss: 1.2628 - val_accuracy: 0.5879\n",
            "\n",
            "Epoch 00001: val_loss improved from inf to 1.26276, saving model to my_best_model.epoch01-loss1.26.hdf5\n",
            "Epoch 2/5\n",
            "1563/1563 [==============================] - 30s 19ms/step - loss: 0.5238 - accuracy: 0.8278 - val_loss: 1.6076 - val_accuracy: 0.5854\n",
            "\n",
            "Epoch 00002: val_loss did not improve from 1.26276\n",
            "Epoch 3/5\n",
            " 352/1563 [=====>........................] - ETA: 22s - loss: 0.3021 - accuracy: 0.9010"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V-fT4cH1PxNe"
      },
      "source": [
        "ModelCheckpoint\n",
        "\n",
        "\n",
        "def inference(image,model):\n",
        "\n",
        "script->\n",
        "python inference.py --input image.jpg\n",
        "CAR"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5-OdMsmxPxIs"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HOTu0EciPL20"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}