Copy_of_Process_GTSRB_nb.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 2
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython2",
      "version": "2.7.6"
    },
    "pycharm": {
      "stem_cell": {
        "cell_type": "raw",
        "source": [],
        "metadata": {
          "collapsed": false
        }
      }
    },
    "colab": {
      "name": "Copy of Process_GTSRB_nb.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "accelerator": "TPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/suchendra-h/stockpredictionai/blob/master/Copy_of_Process_GTSRB_nb.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yxmjmxzWglA9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!pip uninstall tensorflow\n",
        "!pip install tensorflow==2.1\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e6pNBCuiYh3s",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow as tf\n",
        "print(tf.__version__)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "am0KZB8gXbeK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from __future__ import absolute_import, division, print_function, unicode_literals\n",
        "from PIL import Image\n",
        "import os\n",
        "from tensorflow.keras import datasets, layers, models\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, Conv2D, Flatten, Dropout, MaxPooling2D\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "import numpy as np\n",
        "import tensorflow_hub as hub\n",
        "import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "import zipfile, csv, random, pathlib, time\n",
        "from google.colab import drive\n",
        "\n",
        "################################################################################\n",
        "drive.mount('/content/drive')\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "RWSJpsyKqHjH",
        "outputId": "75ddf00a-2ff9-4ccd-d357-8715295902aa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        }
      },
      "source": [
        "### prepare class lables from training and validation folders\n",
        "\n",
        "training_data_dir = \"/content/drive/My Drive/Colab Notebooks/dataset/training\"\n",
        "training_data_dir = pathlib.Path(training_data_dir)\n",
        "image_count = len(list(training_data_dir.glob('*/*.jpg')))\n",
        "tmp = np.array([item.name for item in training_data_dir.glob('*') if item.name != \"LICENSE.txt\"])\n",
        "tmp = np.sort(tmp.astype(int))\n",
        "training_class_names = np.array(['0000' + np.str(i) for i in tmp if i < 10] +\\\n",
        "['000' + np.str(i) for i in tmp if i >= 10])\n",
        "\n",
        "### validation\n",
        "tmp_validation_data_dir = \"/content/drive/My Drive/Colab Notebooks/dataset/validation\"\n",
        "tmp_validation_data_dir = pathlib.Path(tmp_validation_data_dir)\n",
        "tmp = np.array([item.name for item in tmp_validation_data_dir.glob('*') if item.name != \"GT-final_test.csv\"])\n",
        "tmp = np.sort(tmp.astype(int))\n",
        "validation_class_names = np.array(['0000' + np.str(i) for i in tmp if i < 10] +\\\n",
        "['000' + np.str(i) for i in tmp if i >= 10])\n",
        "\n",
        "### prepare call back path\n",
        "checkpoint_path = \"/content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\"\n",
        "checkpoint_dir = os.path.dirname(checkpoint_path)\n",
        "\n",
        "# Create a callback that saves the model's weights every epoch\n",
        "cp_callback = tf.keras.callbacks.ModelCheckpoint(\n",
        "    filepath=checkpoint_path, \n",
        "    verbose=1, \n",
        "    save_weights_only=True,\n",
        "    period=1)"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:`period` argument is deprecated. Please use `save_freq` to specify the frequency in number of samples seen.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "uaIEAptmVGbd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "ff01a9fa-70a6-4f43-9b38-16d1ed639872"
      },
      "source": [
        "start = time.time()\n",
        "# The 1./255 is to convert from uint8 to float32 in range [0,1].\n",
        "image_generator = tf.keras.preprocessing.image.ImageDataGenerator(rescale=1./255)\n",
        "# image_gen_train = ImageDataGenerator(\n",
        "#                     rescale=1./255,\n",
        "#                     rotation_range=45,\n",
        "#                     width_shift_range=.15,\n",
        "#                     height_shift_range=.15,\n",
        "#                     horizontal_flip=True,\n",
        "#                     zoom_range=0.5\n",
        "#                     )\n",
        "validation_image_generator = ImageDataGenerator(rescale=1./255)\n",
        "BATCH_SIZE = 32\n",
        "IMG_HEIGHT = 224\n",
        "IMG_WIDTH = 224\n",
        "STEPS_PER_EPOCH = np.ceil(image_count/BATCH_SIZE)\n",
        "train_data_gen = image_generator.flow_from_directory(directory=str(training_data_dir),\n",
        "                                                     batch_size=BATCH_SIZE,\n",
        "                                                     shuffle=True,\n",
        "                                                     target_size=(IMG_HEIGHT, IMG_WIDTH),\n",
        "                                                     classes = list(training_class_names))\n",
        "# train_data_gen = image_gen_train.flow_from_directory(directory=str(training_data_dir),\n",
        "#                                                      batch_size=BATCH_SIZE,\n",
        "#                                                      shuffle=True,\n",
        "#                                                      target_size=(IMG_HEIGHT, IMG_WIDTH),\n",
        "#                                                      classes = list(training_class_names))\n",
        "validation_gen = validation_image_generator.flow_from_directory(directory=str(tmp_validation_data_dir),\n",
        "                                                     batch_size=BATCH_SIZE,\n",
        "                                                     shuffle=True   ,\n",
        "                                                     target_size=(IMG_HEIGHT, IMG_WIDTH),\n",
        "                                                     classes = list(validation_class_names))\n",
        "epochs = 30\n",
        "model = models.Sequential()\n",
        "model.add(layers.Conv2D(32, 3, padding='same', activation='relu', input_shape=(IMG_HEIGHT, IMG_WIDTH, 3)))\n",
        "model.add(layers.MaxPooling2D())\n",
        "model.add(layers.SpatialDropout2D(0.1))\n",
        "#model.add(layers.Conv2D(32, 3, padding='same', activation='relu'))\n",
        "#model.add(layers.MaxPooling2D())\n",
        "#model.add(layers.SpatialDropout2D(0.1))\n",
        "#model.add(layers.Conv2D(64, 3, padding='same',activation='relu'))\n",
        "#model.add(layers.MaxPooling2D())\n",
        "#model.add(layers.SpatialDropout2D(0.1))\n",
        "#model.add(layers.UpSampling2D())\n",
        "#model.add(layers.Conv2D(128, 3, padding='same', activation='relu'))\n",
        "#model.add(layers.MaxPooling2D())\n",
        "#model.add(layers.SpatialDropout2D(0.1))\n",
        "model.add(Flatten())\n",
        "#model.add(Dense(512, activation='relu'))\n",
        "model.add(Dense(43))\n",
        "\n",
        "model.compile(optimizer='adam',\n",
        "              loss=tf.keras.losses.CategoricalCrossentropy(from_logits=True),\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "model.build(input_shape=(IMG_HEIGHT, IMG_WIDTH, 3))\n",
        "model.summary()\n",
        "# history = model.fit(labeled_training_ds, epochs=5, validation_data=(validation_images, validation_labels))\n",
        "history = model.fit(train_data_gen, epochs=epochs,\n",
        "                    validation_data=validation_gen,\n",
        "                    callbacks=[cp_callback])\n",
        "# save the model\n",
        "saved_model_path = \"/content/drive/My Drive/Colab Notebooks/dataset/saved_models\"\n",
        "model.save('saved_model/my_model')\n",
        "\n",
        "\n",
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "\n",
        "loss=history.history['loss']\n",
        "val_loss=history.history['val_loss']\n",
        "\n",
        "epochs_range = range(epochs)\n",
        "\n",
        "end = time.time()\n",
        "elapsed = end - start\n",
        "print('elapesed time:  ', elapsed)\n",
        "plt.figure(figsize=(8, 8))\n",
        "plt.subplot(1, 2, 1)\n",
        "plt.plot(epochs_range, acc, label='Training Accuracy')\n",
        "plt.plot(epochs_range, val_acc, label='Validation Accuracy')\n",
        "plt.legend(loc='lower right')\n",
        "plt.title('Training and Validation Accuracy')\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "plt.plot(epochs_range, loss, label='Training Loss')\n",
        "plt.plot(epochs_range, val_loss, label='Validation Loss')\n",
        "plt.legend(loc='upper right')\n",
        "plt.title('Training and Validation Loss')\n",
        "plt.show()\n"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 26640 images belonging to 43 classes.\n",
            "Found 12630 images belonging to 43 classes.\n",
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 224, 224, 32)      896       \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 112, 112, 32)      0         \n",
            "_________________________________________________________________\n",
            "spatial_dropout2d (SpatialDr (None, 112, 112, 32)      0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 401408)            0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 43)                17260587  \n",
            "=================================================================\n",
            "Total params: 17,261,483\n",
            "Trainable params: 17,261,483\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "WARNING:tensorflow:sample_weight modes were coerced from\n",
            "  ...\n",
            "    to  \n",
            "  ['...']\n",
            "WARNING:tensorflow:sample_weight modes were coerced from\n",
            "  ...\n",
            "    to  \n",
            "  ['...']\n",
            "Train for 833 steps, validate for 395 steps\n",
            "Epoch 1/30\n",
            "832/833 [============================>.] - ETA: 5s - loss: 1.4490 - accuracy: 0.6968 \n",
            "Epoch 00001: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 6907s 8s/step - loss: 1.4477 - accuracy: 0.6970 - val_loss: 1.1180 - val_accuracy: 0.7127\n",
            "Epoch 2/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.3021 - accuracy: 0.9169\n",
            "Epoch 00002: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1023s 1s/step - loss: 0.3020 - accuracy: 0.9169 - val_loss: 1.1227 - val_accuracy: 0.7579\n",
            "Epoch 3/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.1629 - accuracy: 0.9552\n",
            "Epoch 00003: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1012s 1s/step - loss: 0.1632 - accuracy: 0.9551 - val_loss: 1.2129 - val_accuracy: 0.7543\n",
            "Epoch 4/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.1136 - accuracy: 0.9700\n",
            "Epoch 00004: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1017s 1s/step - loss: 0.1137 - accuracy: 0.9700 - val_loss: 1.2380 - val_accuracy: 0.7735\n",
            "Epoch 5/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0922 - accuracy: 0.9752\n",
            "Epoch 00005: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1019s 1s/step - loss: 0.0922 - accuracy: 0.9752 - val_loss: 1.3747 - val_accuracy: 0.7728\n",
            "Epoch 6/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0743 - accuracy: 0.9812\n",
            "Epoch 00006: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1020s 1s/step - loss: 0.0743 - accuracy: 0.9812 - val_loss: 1.4749 - val_accuracy: 0.7799\n",
            "Epoch 7/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0576 - accuracy: 0.9840\n",
            "Epoch 00007: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1023s 1s/step - loss: 0.0576 - accuracy: 0.9840 - val_loss: 1.6009 - val_accuracy: 0.7717\n",
            "Epoch 8/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0514 - accuracy: 0.9860\n",
            "Epoch 00008: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1026s 1s/step - loss: 0.0515 - accuracy: 0.9860 - val_loss: 1.6739 - val_accuracy: 0.7763\n",
            "Epoch 9/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0407 - accuracy: 0.9893\n",
            "Epoch 00009: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1027s 1s/step - loss: 0.0406 - accuracy: 0.9893 - val_loss: 1.7202 - val_accuracy: 0.7835\n",
            "Epoch 10/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0430 - accuracy: 0.9891\n",
            "Epoch 00010: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1023s 1s/step - loss: 0.0429 - accuracy: 0.9891 - val_loss: 1.9899 - val_accuracy: 0.7829\n",
            "Epoch 11/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0316 - accuracy: 0.9923\n",
            "Epoch 00011: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1019s 1s/step - loss: 0.0316 - accuracy: 0.9923 - val_loss: 2.2281 - val_accuracy: 0.7826\n",
            "Epoch 12/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0356 - accuracy: 0.9909\n",
            "Epoch 00012: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1023s 1s/step - loss: 0.0356 - accuracy: 0.9909 - val_loss: 2.0303 - val_accuracy: 0.7599\n",
            "Epoch 13/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0335 - accuracy: 0.9914\n",
            "Epoch 00013: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1025s 1s/step - loss: 0.0335 - accuracy: 0.9914 - val_loss: 2.1524 - val_accuracy: 0.7793\n",
            "Epoch 14/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0233 - accuracy: 0.9941\n",
            "Epoch 00014: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1027s 1s/step - loss: 0.0233 - accuracy: 0.9941 - val_loss: 2.1426 - val_accuracy: 0.7876\n",
            "Epoch 15/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0234 - accuracy: 0.9943\n",
            "Epoch 00015: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1030s 1s/step - loss: 0.0233 - accuracy: 0.9943 - val_loss: 2.1447 - val_accuracy: 0.7814\n",
            "Epoch 16/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0223 - accuracy: 0.9940\n",
            "Epoch 00016: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1021s 1s/step - loss: 0.0223 - accuracy: 0.9940 - val_loss: 2.3136 - val_accuracy: 0.7769\n",
            "Epoch 17/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0236 - accuracy: 0.9935\n",
            "Epoch 00017: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1018s 1s/step - loss: 0.0236 - accuracy: 0.9935 - val_loss: 2.2969 - val_accuracy: 0.7956\n",
            "Epoch 18/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0281 - accuracy: 0.9936\n",
            "Epoch 00018: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1018s 1s/step - loss: 0.0281 - accuracy: 0.9936 - val_loss: 2.5747 - val_accuracy: 0.7853\n",
            "Epoch 19/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0204 - accuracy: 0.9946\n",
            "Epoch 00019: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1023s 1s/step - loss: 0.0203 - accuracy: 0.9946 - val_loss: 2.9255 - val_accuracy: 0.7894\n",
            "Epoch 20/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0212 - accuracy: 0.9952\n",
            "Epoch 00020: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1021s 1s/step - loss: 0.0212 - accuracy: 0.9952 - val_loss: 2.8493 - val_accuracy: 0.7679\n",
            "Epoch 21/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0146 - accuracy: 0.9965\n",
            "Epoch 00021: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1014s 1s/step - loss: 0.0146 - accuracy: 0.9965 - val_loss: 2.5194 - val_accuracy: 0.7880\n",
            "Epoch 22/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0221 - accuracy: 0.9951\n",
            "Epoch 00022: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1020s 1s/step - loss: 0.0221 - accuracy: 0.9951 - val_loss: 2.5772 - val_accuracy: 0.7966\n",
            "Epoch 23/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0173 - accuracy: 0.9957\n",
            "Epoch 00023: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1018s 1s/step - loss: 0.0173 - accuracy: 0.9957 - val_loss: 2.5388 - val_accuracy: 0.7819\n",
            "Epoch 24/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0188 - accuracy: 0.9962\n",
            "Epoch 00024: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1014s 1s/step - loss: 0.0187 - accuracy: 0.9962 - val_loss: 2.4658 - val_accuracy: 0.8070\n",
            "Epoch 25/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0095 - accuracy: 0.9977\n",
            "Epoch 00025: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1014s 1s/step - loss: 0.0095 - accuracy: 0.9977 - val_loss: 3.0966 - val_accuracy: 0.7753\n",
            "Epoch 26/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0219 - accuracy: 0.9948\n",
            "Epoch 00026: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1013s 1s/step - loss: 0.0219 - accuracy: 0.9948 - val_loss: 2.9427 - val_accuracy: 0.7964\n",
            "Epoch 27/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0106 - accuracy: 0.9975\n",
            "Epoch 00027: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1013s 1s/step - loss: 0.0106 - accuracy: 0.9975 - val_loss: 2.9578 - val_accuracy: 0.7937\n",
            "Epoch 28/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0181 - accuracy: 0.9957\n",
            "Epoch 00028: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1012s 1s/step - loss: 0.0185 - accuracy: 0.9956 - val_loss: 2.9605 - val_accuracy: 0.7524\n",
            "Epoch 29/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0104 - accuracy: 0.9971\n",
            "Epoch 00029: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1013s 1s/step - loss: 0.0104 - accuracy: 0.9971 - val_loss: 2.8194 - val_accuracy: 0.7953\n",
            "Epoch 30/30\n",
            "832/833 [============================>.] - ETA: 1s - loss: 0.0186 - accuracy: 0.9961\n",
            "Epoch 00030: saving model to /content/drive/My Drive/Colab Notebooks/dataset/checkpoints/\n",
            "833/833 [==============================] - 1011s 1s/step - loss: 0.0186 - accuracy: 0.9961 - val_loss: 2.9379 - val_accuracy: 0.7906\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/resource_variable_ops.py:1786: 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: saved_model/my_model/assets\n",
            "elapesed time:   36480.504502534866\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAHiCAYAAADF4pQuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 576x576 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8MV_Kv3BVQ9G",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "cp_callback."
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mgsSmbLPxclT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "outputId": "5ec47bb8-b919-4a15-af91-32f4bfd3004b"
      },
      "source": [
        "validation_class_names\n"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['00000', '00001', '00002', '00003', '00004', '00005', '00006',\n",
              "       '00007', '00008', '00009', '00010', '00011', '00012', '00013',\n",
              "       '00014', '00015', '00016', '00017', '00018', '00019', '00020',\n",
              "       '00021', '00022', '00023', '00024', '00025', '00026', '00027',\n",
              "       '00028', '00029', '00030', '00031', '00032', '00033', '00034',\n",
              "       '00035', '00036', '00037', '00038', '00039', '00040', '00041',\n",
              "       '00042'], dtype='<U5')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eWOVE_Txy23M",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "outputId": "fd040600-bd53-4d36-a4e7-a3cb117d0cf4"
      },
      "source": [
        "training_class_names"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['00000', '00001', '00002', '00003', '00004', '00005', '00006',\n",
              "       '00007', '00008', '00009', '00010', '00011', '00012', '00013',\n",
              "       '00014', '00015', '00016', '00017', '00018', '00019', '00020',\n",
              "       '00021', '00022', '00023', '00024', '00025', '00026', '00027',\n",
              "       '00028', '00029', '00030', '00031', '00032', '00033', '00034',\n",
              "       '00035', '00036', '00037', '00038', '00039', '00040', '00041',\n",
              "       '00042'], dtype='<U5')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sgxQxzJ6y4oh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "ad65b6b1-f123-48d4-c7e1-f7989f229d04"
      },
      "source": [
        "history.history"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'accuracy': [0.6317568],\n",
              " 'loss': [4.542879317311553],\n",
              " 'val_accuracy': [0.67790973],\n",
              " 'val_loss': [1.2747363804261895]}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6CCBpm56Rdny",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        },
        "outputId": "cb1ab572-28c5-4ec4-9ea5-c47ed323543a"
      },
      "source": [
        "model.evaluate(x=validation_gen)"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:sample_weight modes were coerced from\n",
            "  ...\n",
            "    to  \n",
            "  ['...']\n",
            "395/395 [==============================] - 177s 449ms/step - loss: 1.2747 - accuracy: 0.6779\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.2747363804261895, 0.67790973]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t6dSggCIUkCC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "2aac38ea-fa15-4846-8715-644d6a2931b9"
      },
      "source": [
        "validation_gen.classes\n"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([], dtype=int32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a4-oGtgiUzsb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}