farazist.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "farazist.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "z05foMum3swA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!nvidia-smi"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dy5AReIwbuJl",
        "colab_type": "code",
        "outputId": "c1b21ae0-1b33-4c4b-aa18-74a19e11c6f0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive/')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lMlf2cMntzwQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "from tensorflow.keras.models import Sequential, Model\n",
        "from tensorflow.keras.layers import Dense, Flatten\n",
        "from tensorflow.keras.optimizers import SGD, Adam\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "import tensorflow as tf\n",
        "import keras\n",
        "from matplotlib import pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-jqQBbWTBUo9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "directory = \"/content/drive/My Drive/Colab Notebooks/bottle\"\n",
        "width = height = 224\n",
        "batch_size = 32"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S3Ht9aHIbrSg",
        "colab_type": "code",
        "outputId": "5a04fef2-9091-4961-ab4f-aaaeae1ac281",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "data_generator = ImageDataGenerator(\n",
        "    rescale=1./255,\n",
        "    rotation_range=8,\n",
        "    width_shift_range=0.5,\n",
        "    height_shift_range=0.1,\n",
        "    horizontal_flip=True,\n",
        "    vertical_flip=True,\n",
        "    #   zoom = 20,\n",
        "    fill_mode='nearest',\n",
        "    validation_split = 0.1\n",
        "    )\n",
        "\n",
        "train_data = data_generator.flow_from_directory(\n",
        "        directory,\n",
        "        target_size=(width, height),\n",
        "        batch_size = batch_size,\n",
        "        class_mode='categorical',\n",
        "        subset = 'training')\n",
        " \n",
        "validation_data = data_generator.flow_from_directory(\n",
        "        directory,\n",
        "        target_size=(width, height),\n",
        "        class_mode='categorical',\n",
        "        batch_size = batch_size,\n",
        "        shuffle=True,\n",
        "        subset = 'validation')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 2209 images belonging to 28 classes.\n",
            "Found 233 images belonging to 28 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-90GuXSJcf5H",
        "colab_type": "code",
        "outputId": "8a82f534-5d7f-4055-a3ce-376dba7f6765",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "base_model = tf.keras.applications.MobileNetV2(\n",
        "    input_shape=(width, height, 3),\n",
        "    alpha=1.0,\n",
        "    include_top=False, \n",
        "    weights='imagenet', \n",
        "    input_tensor=None, \n",
        "    pooling='avg')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet_v2/mobilenet_v2_weights_tf_dim_ordering_tf_kernels_1.0_224_no_top.h5\n",
            "9412608/9406464 [==============================] - 0s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Evt6EsFhBrHZ",
        "colab_type": "code",
        "cellView": "form",
        "colab": {}
      },
      "source": [
        "#@title\n",
        "base_model = tf.keras.applications.vgg19.VGG19(\n",
        "    input_shape=(width, height, 3),\n",
        "    include_top=False, \n",
        "    weights='imagenet', \n",
        "    input_tensor=None, \n",
        "    pooling='avg')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5M-e6NhdREQz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "print(base_model.summary())"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CwHwloOB1Nba",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tf.keras.utils.plot_model(base_model, show_shapes=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iWL5i6dVwu_W",
        "colab_type": "code",
        "cellView": "form",
        "colab": {}
      },
      "source": [
        "#@title\n",
        "model = Sequential()\n",
        "model.add(base_model)\n",
        "model.add(Dense(4096, activation='relu'))\n",
        "model.add(Dense(1000, activation='relu'))\n",
        "model.add(Dense(28, activation='softmax'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M07Rgn9FdUBo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = tf.keras.models.Sequential([\n",
        "    base_model,\n",
        "    Dense(28, activation='softmax')\n",
        "])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "seoZekgdTq5a",
        "colab_type": "code",
        "cellView": "form",
        "colab": {}
      },
      "source": [
        "#@title\n",
        "model = tf.keras.models.Sequential([\n",
        "    base_model,\n",
        "    Dense(4096, activation='relu'),\n",
        "    Dense(1000, activation='relu'),\n",
        "    Dense(28, activation='softmax')\n",
        "])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JYvEpFeRdtGg",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(loss=tf.keras.losses.categorical_crossentropy,\n",
        "              optimizer=tf.keras.optimizers.Adadelta(\n",
        "                  learning_rate=0.001, \n",
        "                  rho=0.95, \n",
        "                  epsilon=1e-07, \n",
        "                  name='Adadelta'),\n",
        "              metrics=['acc'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QAUV1CMidzKU",
        "colab_type": "code",
        "outputId": "bf4defa8-1311-41f8-e87c-da1ae9c1e75a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "history = model.fit(\n",
        "    train_data,\n",
        "    steps_per_epoch = train_data.samples // batch_size,\n",
        "    epochs=100,\n",
        "    shuffle = True,\n",
        "    validation_data=validation_data,\n",
        "    validation_steps = validation_data.samples // batch_size,\n",
        "    verbose = 1)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "69/69 [==============================] - 763s 11s/step - loss: 3.5640 - acc: 0.0473 - val_loss: 3.6617 - val_acc: 0.0312\n",
            "Epoch 2/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 3.4634 - acc: 0.0634 - val_loss: 3.6477 - val_acc: 0.0491\n",
            "Epoch 3/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 3.3933 - acc: 0.0689 - val_loss: 3.5791 - val_acc: 0.0446\n",
            "Epoch 4/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 3.3068 - acc: 0.0868 - val_loss: 3.5193 - val_acc: 0.0446\n",
            "Epoch 5/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 3.2451 - acc: 0.0988 - val_loss: 3.4727 - val_acc: 0.0580\n",
            "Epoch 6/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 3.1646 - acc: 0.1203 - val_loss: 3.4099 - val_acc: 0.0491\n",
            "Epoch 7/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 3.1170 - acc: 0.1364 - val_loss: 3.3950 - val_acc: 0.0714\n",
            "Epoch 8/100\n",
            "69/69 [==============================] - 87s 1s/step - loss: 3.0341 - acc: 0.1612 - val_loss: 3.2734 - val_acc: 0.0938\n",
            "Epoch 9/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 2.9761 - acc: 0.1750 - val_loss: 3.3137 - val_acc: 0.0938\n",
            "Epoch 10/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 2.9226 - acc: 0.2017 - val_loss: 3.1819 - val_acc: 0.1161\n",
            "Epoch 11/100\n",
            "69/69 [==============================] - 81s 1s/step - loss: 2.8462 - acc: 0.2352 - val_loss: 3.1737 - val_acc: 0.1161\n",
            "Epoch 12/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 2.7780 - acc: 0.2549 - val_loss: 3.1110 - val_acc: 0.1116\n",
            "Epoch 13/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 2.7129 - acc: 0.2751 - val_loss: 3.1381 - val_acc: 0.1339\n",
            "Epoch 14/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 2.6728 - acc: 0.2825 - val_loss: 3.0052 - val_acc: 0.1607\n",
            "Epoch 15/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 2.6130 - acc: 0.3101 - val_loss: 3.0221 - val_acc: 0.1429\n",
            "Epoch 16/100\n",
            "69/69 [==============================] - 81s 1s/step - loss: 2.5476 - acc: 0.3225 - val_loss: 2.8957 - val_acc: 0.1920\n",
            "Epoch 17/100\n",
            "69/69 [==============================] - 83s 1s/step - loss: 2.4786 - acc: 0.3505 - val_loss: 2.8909 - val_acc: 0.2098\n",
            "Epoch 18/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 2.4538 - acc: 0.3638 - val_loss: 2.8548 - val_acc: 0.1964\n",
            "Epoch 19/100\n",
            "69/69 [==============================] - 83s 1s/step - loss: 2.3841 - acc: 0.3863 - val_loss: 2.7897 - val_acc: 0.2321\n",
            "Epoch 20/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 2.3209 - acc: 0.4093 - val_loss: 2.7423 - val_acc: 0.2054\n",
            "Epoch 21/100\n",
            "69/69 [==============================] - 81s 1s/step - loss: 2.2779 - acc: 0.4231 - val_loss: 2.7222 - val_acc: 0.2054\n",
            "Epoch 22/100\n",
            "69/69 [==============================] - 83s 1s/step - loss: 2.2310 - acc: 0.4419 - val_loss: 2.5617 - val_acc: 0.2723\n",
            "Epoch 23/100\n",
            "69/69 [==============================] - 83s 1s/step - loss: 2.1765 - acc: 0.4557 - val_loss: 2.6116 - val_acc: 0.2723\n",
            "Epoch 24/100\n",
            "69/69 [==============================] - 87s 1s/step - loss: 2.1317 - acc: 0.4786 - val_loss: 2.5135 - val_acc: 0.3170\n",
            "Epoch 25/100\n",
            "69/69 [==============================] - 89s 1s/step - loss: 2.0793 - acc: 0.4851 - val_loss: 2.4164 - val_acc: 0.3348\n",
            "Epoch 26/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 2.0407 - acc: 0.4906 - val_loss: 2.4123 - val_acc: 0.3750\n",
            "Epoch 27/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.9698 - acc: 0.5305 - val_loss: 2.2978 - val_acc: 0.3839\n",
            "Epoch 28/100\n",
            "69/69 [==============================] - 87s 1s/step - loss: 1.9624 - acc: 0.5250 - val_loss: 2.2747 - val_acc: 0.4241\n",
            "Epoch 29/100\n",
            "69/69 [==============================] - 87s 1s/step - loss: 1.8876 - acc: 0.5521 - val_loss: 2.2121 - val_acc: 0.4643\n",
            "Epoch 30/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.8425 - acc: 0.5613 - val_loss: 2.2649 - val_acc: 0.4018\n",
            "Epoch 31/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.8381 - acc: 0.5599 - val_loss: 2.0084 - val_acc: 0.5357\n",
            "Epoch 32/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.7643 - acc: 0.5898 - val_loss: 2.0602 - val_acc: 0.4955\n",
            "Epoch 33/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.7414 - acc: 0.6059 - val_loss: 2.0141 - val_acc: 0.5000\n",
            "Epoch 34/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.6952 - acc: 0.6123 - val_loss: 1.8748 - val_acc: 0.5536\n",
            "Epoch 35/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 1.6500 - acc: 0.6169 - val_loss: 1.8759 - val_acc: 0.5670\n",
            "Epoch 36/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.6261 - acc: 0.6403 - val_loss: 1.7887 - val_acc: 0.5938\n",
            "Epoch 37/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 1.5855 - acc: 0.6426 - val_loss: 1.7909 - val_acc: 0.5893\n",
            "Epoch 38/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 1.5604 - acc: 0.6449 - val_loss: 1.6287 - val_acc: 0.6562\n",
            "Epoch 39/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 1.5088 - acc: 0.6808 - val_loss: 1.7189 - val_acc: 0.6205\n",
            "Epoch 40/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 1.4655 - acc: 0.6909 - val_loss: 1.6375 - val_acc: 0.5982\n",
            "Epoch 41/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 1.4424 - acc: 0.6936 - val_loss: 1.5617 - val_acc: 0.7098\n",
            "Epoch 42/100\n",
            "69/69 [==============================] - 79s 1s/step - loss: 1.4146 - acc: 0.6936 - val_loss: 1.4539 - val_acc: 0.7277\n",
            "Epoch 43/100\n",
            "69/69 [==============================] - 80s 1s/step - loss: 1.3668 - acc: 0.7074 - val_loss: 1.4904 - val_acc: 0.6830\n",
            "Epoch 44/100\n",
            "69/69 [==============================] - 79s 1s/step - loss: 1.3630 - acc: 0.7166 - val_loss: 1.4465 - val_acc: 0.6786\n",
            "Epoch 45/100\n",
            "69/69 [==============================] - 78s 1s/step - loss: 1.3061 - acc: 0.7400 - val_loss: 1.3843 - val_acc: 0.7098\n",
            "Epoch 46/100\n",
            "69/69 [==============================] - 77s 1s/step - loss: 1.2903 - acc: 0.7115 - val_loss: 1.3315 - val_acc: 0.7366\n",
            "Epoch 47/100\n",
            "69/69 [==============================] - 78s 1s/step - loss: 1.2411 - acc: 0.7565 - val_loss: 1.3052 - val_acc: 0.7679\n",
            "Epoch 48/100\n",
            "69/69 [==============================] - 77s 1s/step - loss: 1.2045 - acc: 0.7735 - val_loss: 1.2822 - val_acc: 0.7589\n",
            "Epoch 49/100\n",
            "69/69 [==============================] - 80s 1s/step - loss: 1.1920 - acc: 0.7588 - val_loss: 1.1915 - val_acc: 0.7857\n",
            "Epoch 50/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 1.1637 - acc: 0.7772 - val_loss: 1.1646 - val_acc: 0.7634\n",
            "Epoch 51/100\n",
            "69/69 [==============================] - 93s 1s/step - loss: 1.1494 - acc: 0.7827 - val_loss: 1.1530 - val_acc: 0.7768\n",
            "Epoch 52/100\n",
            "69/69 [==============================] - 87s 1s/step - loss: 1.1215 - acc: 0.7827 - val_loss: 1.2085 - val_acc: 0.7455\n",
            "Epoch 53/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 1.0916 - acc: 0.7933 - val_loss: 1.0888 - val_acc: 0.7768\n",
            "Epoch 54/100\n",
            "69/69 [==============================] - 87s 1s/step - loss: 1.0652 - acc: 0.8140 - val_loss: 1.0266 - val_acc: 0.8259\n",
            "Epoch 55/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.0431 - acc: 0.8052 - val_loss: 1.0271 - val_acc: 0.8259\n",
            "Epoch 56/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 1.0193 - acc: 0.8195 - val_loss: 0.9868 - val_acc: 0.8170\n",
            "Epoch 57/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.9930 - acc: 0.8121 - val_loss: 1.0064 - val_acc: 0.7991\n",
            "Epoch 58/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.9772 - acc: 0.8277 - val_loss: 0.9933 - val_acc: 0.8304\n",
            "Epoch 59/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.9760 - acc: 0.8232 - val_loss: 0.9561 - val_acc: 0.8393\n",
            "Epoch 60/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 0.9224 - acc: 0.8369 - val_loss: 0.9172 - val_acc: 0.8259\n",
            "Epoch 61/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 0.9159 - acc: 0.8415 - val_loss: 0.8675 - val_acc: 0.8705\n",
            "Epoch 62/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.8959 - acc: 0.8424 - val_loss: 0.9179 - val_acc: 0.8393\n",
            "Epoch 63/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.8749 - acc: 0.8438 - val_loss: 0.8431 - val_acc: 0.8393\n",
            "Epoch 64/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 0.8628 - acc: 0.8434 - val_loss: 0.7843 - val_acc: 0.8527\n",
            "Epoch 65/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.8247 - acc: 0.8590 - val_loss: 0.8356 - val_acc: 0.8304\n",
            "Epoch 66/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.8293 - acc: 0.8503 - val_loss: 0.7652 - val_acc: 0.8839\n",
            "Epoch 67/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.8019 - acc: 0.8668 - val_loss: 0.7251 - val_acc: 0.8929\n",
            "Epoch 68/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.7846 - acc: 0.8654 - val_loss: 0.7241 - val_acc: 0.8929\n",
            "Epoch 69/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.7807 - acc: 0.8691 - val_loss: 0.8272 - val_acc: 0.8438\n",
            "Epoch 70/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.7546 - acc: 0.8695 - val_loss: 0.7136 - val_acc: 0.8616\n",
            "Epoch 71/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.7240 - acc: 0.8792 - val_loss: 0.7364 - val_acc: 0.8839\n",
            "Epoch 72/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.7395 - acc: 0.8682 - val_loss: 0.6729 - val_acc: 0.8929\n",
            "Epoch 73/100\n",
            "69/69 [==============================] - 89s 1s/step - loss: 0.7172 - acc: 0.8769 - val_loss: 0.6911 - val_acc: 0.8839\n",
            "Epoch 74/100\n",
            "69/69 [==============================] - 87s 1s/step - loss: 0.7082 - acc: 0.8806 - val_loss: 0.6010 - val_acc: 0.9196\n",
            "Epoch 75/100\n",
            "69/69 [==============================] - 88s 1s/step - loss: 0.6901 - acc: 0.8801 - val_loss: 0.6175 - val_acc: 0.9062\n",
            "Epoch 76/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.6788 - acc: 0.8930 - val_loss: 0.6338 - val_acc: 0.8973\n",
            "Epoch 77/100\n",
            "69/69 [==============================] - 83s 1s/step - loss: 0.6499 - acc: 0.8921 - val_loss: 0.5826 - val_acc: 0.9196\n",
            "Epoch 78/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 0.6649 - acc: 0.8934 - val_loss: 0.5680 - val_acc: 0.9107\n",
            "Epoch 79/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 0.6473 - acc: 0.8957 - val_loss: 0.4983 - val_acc: 0.9330\n",
            "Epoch 80/100\n",
            "69/69 [==============================] - 82s 1s/step - loss: 0.6125 - acc: 0.9022 - val_loss: 0.5802 - val_acc: 0.8929\n",
            "Epoch 81/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 0.6223 - acc: 0.8916 - val_loss: 0.5617 - val_acc: 0.9152\n",
            "Epoch 82/100\n",
            "69/69 [==============================] - 83s 1s/step - loss: 0.6163 - acc: 0.8981 - val_loss: 0.5151 - val_acc: 0.9196\n",
            "Epoch 83/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 0.5925 - acc: 0.9003 - val_loss: 0.5688 - val_acc: 0.9152\n",
            "Epoch 84/100\n",
            "69/69 [==============================] - 84s 1s/step - loss: 0.5754 - acc: 0.9104 - val_loss: 0.4530 - val_acc: 0.9330\n",
            "Epoch 85/100\n",
            "69/69 [==============================] - 88s 1s/step - loss: 0.5838 - acc: 0.9008 - val_loss: 0.5640 - val_acc: 0.8973\n",
            "Epoch 86/100\n",
            "69/69 [==============================] - 90s 1s/step - loss: 0.5683 - acc: 0.9035 - val_loss: 0.5355 - val_acc: 0.9241\n",
            "Epoch 87/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.5558 - acc: 0.9072 - val_loss: 0.5050 - val_acc: 0.9241\n",
            "Epoch 88/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.5266 - acc: 0.9123 - val_loss: 0.5164 - val_acc: 0.9375\n",
            "Epoch 89/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.5220 - acc: 0.9159 - val_loss: 0.4898 - val_acc: 0.9286\n",
            "Epoch 90/100\n",
            "69/69 [==============================] - 89s 1s/step - loss: 0.5324 - acc: 0.9063 - val_loss: 0.4186 - val_acc: 0.9509\n",
            "Epoch 91/100\n",
            "69/69 [==============================] - 87s 1s/step - loss: 0.5264 - acc: 0.9068 - val_loss: 0.4412 - val_acc: 0.9241\n",
            "Epoch 92/100\n",
            "69/69 [==============================] - 88s 1s/step - loss: 0.4906 - acc: 0.9178 - val_loss: 0.4554 - val_acc: 0.9196\n",
            "Epoch 93/100\n",
            "69/69 [==============================] - 88s 1s/step - loss: 0.5105 - acc: 0.9127 - val_loss: 0.4346 - val_acc: 0.9330\n",
            "Epoch 94/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.4868 - acc: 0.9136 - val_loss: 0.3919 - val_acc: 0.9509\n",
            "Epoch 95/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.4776 - acc: 0.9201 - val_loss: 0.3899 - val_acc: 0.9598\n",
            "Epoch 96/100\n",
            "69/69 [==============================] - 85s 1s/step - loss: 0.4817 - acc: 0.9173 - val_loss: 0.4735 - val_acc: 0.9286\n",
            "Epoch 97/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.4807 - acc: 0.9123 - val_loss: 0.4705 - val_acc: 0.9062\n",
            "Epoch 98/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.4672 - acc: 0.9219 - val_loss: 0.3892 - val_acc: 0.9554\n",
            "Epoch 99/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.4607 - acc: 0.9150 - val_loss: 0.4427 - val_acc: 0.9196\n",
            "Epoch 100/100\n",
            "69/69 [==============================] - 86s 1s/step - loss: 0.4561 - acc: 0.9302 - val_loss: 0.4013 - val_acc: 0.9375\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BM-3ywe3YZKb",
        "colab_type": "code",
        "outputId": "b4389f69-b614-4733-b92e-79af64013738",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "source": [
        "from matplotlib import pyplot as plt\n",
        "\n",
        "# plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['acc'])\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sCSOa5gg29lJ",
        "colab_type": "code",
        "outputId": "492546ab-e52a-4866-fe6f-18603694fc02",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 105
        }
      },
      "source": [
        "model.save('my_model')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/resource_variable_ops.py:1817: calling BaseResourceVariable.__init__ (from tensorflow.python.ops.resource_variable_ops) with constraint is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "If using Keras pass *_constraint arguments to layers.\n",
            "INFO:tensorflow:Assets written to: my_model/assets\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bAP679oKpdjl",
        "colab_type": "code",
        "outputId": "ebe4e7ae-f12b-4365-a0f4-d7b3b845a95c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import shutil\n",
        "shutil.make_archive('my_model', 'zip', 'my_model')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'/content/my_model.zip.zip'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gNhRaY8etsSK",
        "colab_type": "code",
        "outputId": "f041a9a0-9e8b-4a73-9453-2a4e904037c5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import tensorflow as tf\n",
        "\n",
        "converter = tf.lite.TFLiteConverter.from_saved_model('my_model')\n",
        "tflite_model = converter.convert()\n",
        "open(\"farazist.tflite\", \"wb\").write(tflite_model)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "8992236"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o5B8CA065oaU",
        "colab_type": "code",
        "outputId": "dca5c664-7046-4438-cdd5-23e9cf9daa30",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        }
      },
      "source": [
        "import cv2 as cv\n",
        "img = cv.imread('photo.jpg')\n",
        "img = cv.resize(img, (299, 299))\n",
        "img = cv.cvtColor(img, cv.COLOR_BGR2RGB)\n",
        "\n",
        "img = img / 255.0\n",
        "\n",
        "img = img.reshape(1, 299, 299, 3)\n",
        "\n",
        "y_pred = model.predict(img)\n",
        "print(y_pred)\n",
        "print(np.argmax(y_pred))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[2.9779755e-04 4.9298689e-07 7.6740027e-05 1.3062366e-09 7.1239114e-08\n",
            "  2.4361536e-02 9.1198422e-03 1.0068750e-05 5.9401607e-08 6.7065279e-03\n",
            "  9.4462734e-01 8.7986784e-03 2.7636106e-05 2.2263576e-03 4.7452792e-07\n",
            "  3.4970781e-03 6.0583574e-05 3.6783764e-05 1.3073877e-04 2.1177702e-05]]\n",
            "10\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UQZ5b79ypnCt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RchudczSpow7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from IPython.display import display, Javascript\n",
        "from google.colab.output import eval_js\n",
        "from base64 import b64decode\n",
        "\n",
        "def take_photo(filename='photo.jpg', quality=0.8):\n",
        "  js = Javascript('''\n",
        "    async function takePhoto(quality) {\n",
        "      const div = document.createElement('div');\n",
        "      const capture = document.createElement('button');\n",
        "      capture.textContent = 'Capture';\n",
        "      div.appendChild(capture);\n",
        "\n",
        "      const video = document.createElement('video');\n",
        "      video.style.display = 'block';\n",
        "      const stream = await navigator.mediaDevices.getUserMedia({video: true});\n",
        "\n",
        "      document.body.appendChild(div);\n",
        "      div.appendChild(video);\n",
        "      video.srcObject = stream;\n",
        "      await video.play();\n",
        "\n",
        "      // Resize the output to fit the video element.\n",
        "      google.colab.output.setIframeHeight(document.documentElement.scrollHeight, true);\n",
        "\n",
        "      // Wait for Capture to be clicked.\n",
        "      await new Promise((resolve) => capture.onclick = resolve);\n",
        "\n",
        "      const canvas = document.createElement('canvas');\n",
        "      canvas.width = video.videoWidth;\n",
        "      canvas.height = video.videoHeight;\n",
        "      canvas.getContext('2d').drawImage(video, 0, 0);\n",
        "      stream.getVideoTracks()[0].stop();\n",
        "      div.remove();\n",
        "      return canvas.toDataURL('image/jpeg', quality);\n",
        "    }\n",
        "    ''')\n",
        "  display(js)\n",
        "  data = eval_js('takePhoto({})'.format(quality))\n",
        "  binary = b64decode(data.split(',')[1])\n",
        "  with open(filename, 'wb') as f:\n",
        "    f.write(binary)\n",
        "  return filename"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8uTT-BgHpouq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from IPython.display import Image\n",
        "try:\n",
        "  filename = take_photo()\n",
        "  print('Saved to {}'.format(filename))\n",
        "  \n",
        "  # Show the image which was just taken.\n",
        "  display(Image(filename))\n",
        "except Exception as err:\n",
        "  # Errors will be thrown if the user does not have a webcam or if they do not\n",
        "  # grant the page permission to access it.\n",
        "  print(str(err))"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}