Disease_Detection_(Final).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Disease Detection (Final).ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "dwF7iNXXifvF"
      },
      "source": [
        "from google.colab import drive\r\n",
        "drive.mount('/content/gdrive')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "97w90hXH06TE"
      },
      "source": [
        "Data Pre-Processing and Augmentation using Keras ImageDataGenerator"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9d-GaiXcqz5-"
      },
      "source": [
        "import tensorflow as tf \r\n",
        "\r\n",
        "datagen = tf.keras.preprocessing.image.ImageDataGenerator(rescale = 1./255., \r\n",
        "                                                          rotation_range = 8, \r\n",
        "                                                          width_shift_range = 0.1, \r\n",
        "                                                          height_shift_range = 0.1,\r\n",
        "                                                          shear_range = 0.2, \r\n",
        "                                                          zoom_range = 0.2,\r\n",
        "                                                          brightness_range = [0.5,1.5],\r\n",
        "                                                          horizontal_flip = True,\r\n",
        "                                                          fill_mode = 'nearest',\r\n",
        "                                                          dtype = None,                                                      \r\n",
        "                                                          validation_split = 0.075)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eXBq2dM61VKX"
      },
      "source": [
        "Defining generators for training and validation set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2m_-EOWSq2jk"
      },
      "source": [
        "dire = '/content/gdrive/MyDrive/Eye Diseases'\r\n",
        "\r\n",
        "train_generator = datagen.flow_from_directory(\r\n",
        "    dire,\r\n",
        "    subset=\"training\",\r\n",
        "    batch_size=32,\r\n",
        "    class_mode='categorical',     \r\n",
        "    shuffle=True,                      \r\n",
        "    target_size=(200,200)\r\n",
        ")\r\n",
        "\r\n",
        "valid_generator = datagen.flow_from_directory(\r\n",
        "    dire,\r\n",
        "    subset=\"validation\",\r\n",
        "    shuffle=True,\r\n",
        "    batch_size=32,\r\n",
        "    class_mode='categorical',                           \r\n",
        "    target_size=(200,200)\r\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L3p7y7bprHog"
      },
      "source": [
        "for image_batch, label_batch in train_generator:\r\n",
        "  break\r\n",
        "image_batch.shape, label_batch.shape"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BqUKLVhW1oMU"
      },
      "source": [
        "Deducing and saving labels"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Q5REFgn7rK1C",
        "outputId": "20a65d4f-69ee-4238-927d-3fbdf672a2a2"
      },
      "source": [
        "label_batch[0:5]\r\n",
        "print(train_generator.class_indices)\r\n",
        "labels = '\\n'.join(sorted(train_generator.class_indices.keys()))\r\n",
        "with open('labels.txt', 'w') as f:\r\n",
        " f.write(labels)\r\n",
        "!cat labels.txt"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "{'Arcus Senilis': 0, 'Cataracts': 1, 'Endophthalmitis': 2, 'Hemangioma': 3, 'Hyphema': 4}\n",
            "Arcus Senilis\n",
            "Cataracts\n",
            "Endophthalmitis\n",
            "Hemangioma\n",
            "Hyphema"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gecl_hg31tHU"
      },
      "source": [
        "Retrieving vgg16 pretrained vgg weights"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7B7NwsparOnM",
        "outputId": "d3c98a1e-14b3-4a80-d9dd-3ac867fe2254"
      },
      "source": [
        "import os\r\n",
        "import urllib.request\r\n",
        "\r\n",
        "urllib.request.urlretrieve('https://github.com/fchollet/deep-learning-models/releases/download/v0.1/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5',\r\n",
        "                           'vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5')\r\n",
        "os.listdir()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['.config',\n",
              " 'vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5',\n",
              " 'gdrive',\n",
              " 'labels.txt',\n",
              " 'sample_data']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iya3r2xI2S6u"
      },
      "source": [
        "Initializing vgg model with downloaded, pretrained weights and input shape of 200,200,3 because of the small image size in the dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OstfLUIqrRJr"
      },
      "source": [
        "from tensorflow.keras import layers \r\n",
        "from tensorflow.keras import Model \r\n",
        "\r\n",
        "from tensorflow.keras.applications import ResNet50\r\n",
        "from tensorflow.python.keras.models import Sequential\r\n",
        "from tensorflow.python.keras.layers import Dense, Flatten, GlobalAveragePooling2D\r\n",
        "\r\n",
        "from keras.applications.vgg16 import VGG16, preprocess_input\r\n",
        "\r\n",
        "vgg16_weight_path = '/content/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5'\r\n",
        "\r\n",
        "vgg = VGG16(\r\n",
        "    weights = vgg16_weight_path,\r\n",
        "    include_top = False, \r\n",
        "    input_shape = (200,200,3)\r\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "txa_3Yzt2w7m"
      },
      "source": [
        "Freezing previous layers "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "exh2HGi-rTci"
      },
      "source": [
        "for layer in vgg.layers:\r\n",
        "    layer.trainable = False"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9by8nn9y3kBY"
      },
      "source": [
        "##Adding layers for training \r\n",
        "\r\n",
        "best accuracy: 128 -> 32 -> 5 for final activation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5d8wgh1hrVvY"
      },
      "source": [
        "from keras import layers\r\n",
        "from tensorflow.keras import Sequential\r\n",
        "from tensorflow.keras.layers import Flatten, Dense\r\n",
        "from tensorflow.keras import regularizers\r\n",
        "\r\n",
        "\r\n",
        "\r\n",
        "model = Sequential()\r\n",
        "model.add(vgg)\r\n",
        "model.add(Flatten())\r\n",
        "model.add(Dense(128, activation='relu'))\r\n",
        "model.add(layers.Dropout(rate=0.5))\r\n",
        "model.add(Dense(32, activation='relu'))\r\n",
        "model.add(layers.Dropout(rate=0.2))\r\n",
        "#model.add(Dense(16, activation='relu'))\r\n",
        "#model.add(layers.Dropout(rate=0.1))\r\n",
        "model.add(Dense(5,activation=\"softmax\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1wLm7nN635GQ"
      },
      "source": [
        "Slow learning rate with larger number of epochs has proven to give more accurate results without overfitting"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E7oEmB3LrZKr"
      },
      "source": [
        "opt = tf.keras.optimizers.Adam(learning_rate=0.0001)\r\n",
        "\r\n",
        "model.compile(optimizer=opt, loss=\"categorical_crossentropy\", metrics=[\"accuracy\"])\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kDgI6dw_rbqO",
        "outputId": "6cb88787-57c2-42f9-992b-91a73d72a305"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "vgg16 (Functional)           (None, 6, 6, 512)         14714688  \n",
            "_________________________________________________________________\n",
            "flatten_2 (Flatten)          (None, 18432)             0         \n",
            "_________________________________________________________________\n",
            "dense_7 (Dense)              (None, 128)               2359424   \n",
            "_________________________________________________________________\n",
            "dropout_5 (Dropout)          (None, 128)               0         \n",
            "_________________________________________________________________\n",
            "dense_8 (Dense)              (None, 32)                4128      \n",
            "_________________________________________________________________\n",
            "dropout_6 (Dropout)          (None, 32)                0         \n",
            "_________________________________________________________________\n",
            "dense_9 (Dense)              (None, 5)                 165       \n",
            "=================================================================\n",
            "Total params: 17,078,405\n",
            "Trainable params: 2,363,717\n",
            "Non-trainable params: 14,714,688\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sW9DKzFr4B9m"
      },
      "source": [
        "Training for first 300 epochs"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gv2gLV54rdEr",
        "outputId": "b08da521-1685-4f0a-9989-bba3ef7e7b6d"
      },
      "source": [
        "hist= model.fit(train_generator, validation_data = valid_generator, epochs = 300)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/300\n",
            "33/33 [==============================] - 19s 521ms/step - loss: 1.7537 - accuracy: 0.2769 - val_loss: 1.4515 - val_accuracy: 0.4096\n",
            "Epoch 2/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 1.4878 - accuracy: 0.3669 - val_loss: 1.3629 - val_accuracy: 0.5301\n",
            "Epoch 3/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 1.3833 - accuracy: 0.4446 - val_loss: 1.3419 - val_accuracy: 0.4699\n",
            "Epoch 4/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 1.4016 - accuracy: 0.3915 - val_loss: 1.2581 - val_accuracy: 0.5060\n",
            "Epoch 5/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 1.3530 - accuracy: 0.4132 - val_loss: 1.2836 - val_accuracy: 0.5301\n",
            "Epoch 6/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 1.3432 - accuracy: 0.4480 - val_loss: 1.2664 - val_accuracy: 0.4578\n",
            "Epoch 7/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 1.3387 - accuracy: 0.4464 - val_loss: 1.2287 - val_accuracy: 0.5181\n",
            "Epoch 8/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 1.2763 - accuracy: 0.5094 - val_loss: 1.1894 - val_accuracy: 0.5301\n",
            "Epoch 9/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 1.2209 - accuracy: 0.5284 - val_loss: 1.2016 - val_accuracy: 0.5060\n",
            "Epoch 10/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 1.2476 - accuracy: 0.4879 - val_loss: 1.1254 - val_accuracy: 0.4940\n",
            "Epoch 11/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 1.1673 - accuracy: 0.5117 - val_loss: 1.1314 - val_accuracy: 0.5301\n",
            "Epoch 12/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 1.1606 - accuracy: 0.5340 - val_loss: 1.1086 - val_accuracy: 0.5422\n",
            "Epoch 13/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 1.1594 - accuracy: 0.5319 - val_loss: 1.1018 - val_accuracy: 0.5783\n",
            "Epoch 14/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 1.1363 - accuracy: 0.5493 - val_loss: 1.0635 - val_accuracy: 0.5181\n",
            "Epoch 15/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 1.1791 - accuracy: 0.5334 - val_loss: 1.1125 - val_accuracy: 0.5422\n",
            "Epoch 16/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 1.1155 - accuracy: 0.5344 - val_loss: 1.0690 - val_accuracy: 0.5542\n",
            "Epoch 17/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 1.1234 - accuracy: 0.5427 - val_loss: 1.0530 - val_accuracy: 0.5663\n",
            "Epoch 18/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 1.1126 - accuracy: 0.5369 - val_loss: 1.0166 - val_accuracy: 0.6024\n",
            "Epoch 19/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 1.0869 - accuracy: 0.5554 - val_loss: 0.9872 - val_accuracy: 0.5904\n",
            "Epoch 20/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 1.0013 - accuracy: 0.6033 - val_loss: 1.0237 - val_accuracy: 0.5542\n",
            "Epoch 21/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 1.0273 - accuracy: 0.6084 - val_loss: 1.0091 - val_accuracy: 0.5663\n",
            "Epoch 22/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.9832 - accuracy: 0.6202 - val_loss: 0.9940 - val_accuracy: 0.5904\n",
            "Epoch 23/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 1.0230 - accuracy: 0.5892 - val_loss: 0.9328 - val_accuracy: 0.6145\n",
            "Epoch 24/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 1.0363 - accuracy: 0.6294 - val_loss: 0.9840 - val_accuracy: 0.5663\n",
            "Epoch 25/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.9377 - accuracy: 0.6513 - val_loss: 1.0407 - val_accuracy: 0.5904\n",
            "Epoch 26/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.9699 - accuracy: 0.6300 - val_loss: 0.9755 - val_accuracy: 0.6867\n",
            "Epoch 27/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.9713 - accuracy: 0.6246 - val_loss: 0.9267 - val_accuracy: 0.6506\n",
            "Epoch 28/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.9016 - accuracy: 0.6545 - val_loss: 0.9774 - val_accuracy: 0.5904\n",
            "Epoch 29/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.9329 - accuracy: 0.6396 - val_loss: 0.8959 - val_accuracy: 0.6265\n",
            "Epoch 30/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.9424 - accuracy: 0.6435 - val_loss: 0.9264 - val_accuracy: 0.6627\n",
            "Epoch 31/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.9453 - accuracy: 0.6460 - val_loss: 0.9296 - val_accuracy: 0.6265\n",
            "Epoch 32/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.8996 - accuracy: 0.6614 - val_loss: 0.9141 - val_accuracy: 0.6506\n",
            "Epoch 33/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.8840 - accuracy: 0.6749 - val_loss: 0.8476 - val_accuracy: 0.6627\n",
            "Epoch 34/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.8759 - accuracy: 0.6643 - val_loss: 0.9106 - val_accuracy: 0.6627\n",
            "Epoch 35/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.8451 - accuracy: 0.6976 - val_loss: 0.9405 - val_accuracy: 0.6506\n",
            "Epoch 36/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.8268 - accuracy: 0.6855 - val_loss: 0.8654 - val_accuracy: 0.6867\n",
            "Epoch 37/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.8871 - accuracy: 0.6729 - val_loss: 0.8609 - val_accuracy: 0.7108\n",
            "Epoch 38/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.8016 - accuracy: 0.7142 - val_loss: 0.8634 - val_accuracy: 0.7108\n",
            "Epoch 39/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.8418 - accuracy: 0.6833 - val_loss: 0.9306 - val_accuracy: 0.6265\n",
            "Epoch 40/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.8265 - accuracy: 0.6833 - val_loss: 0.8538 - val_accuracy: 0.6506\n",
            "Epoch 41/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.8350 - accuracy: 0.6920 - val_loss: 0.8875 - val_accuracy: 0.6867\n",
            "Epoch 42/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.8200 - accuracy: 0.7070 - val_loss: 0.9265 - val_accuracy: 0.6145\n",
            "Epoch 43/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.8228 - accuracy: 0.6908 - val_loss: 0.8423 - val_accuracy: 0.6627\n",
            "Epoch 44/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.7954 - accuracy: 0.6757 - val_loss: 0.8518 - val_accuracy: 0.6867\n",
            "Epoch 45/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.7491 - accuracy: 0.7206 - val_loss: 0.7886 - val_accuracy: 0.7108\n",
            "Epoch 46/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.7876 - accuracy: 0.7317 - val_loss: 0.7833 - val_accuracy: 0.7229\n",
            "Epoch 47/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.7705 - accuracy: 0.7088 - val_loss: 0.7871 - val_accuracy: 0.7349\n",
            "Epoch 48/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.7309 - accuracy: 0.7605 - val_loss: 0.8516 - val_accuracy: 0.6506\n",
            "Epoch 49/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.7670 - accuracy: 0.7133 - val_loss: 0.8697 - val_accuracy: 0.6747\n",
            "Epoch 50/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.7574 - accuracy: 0.7039 - val_loss: 0.7387 - val_accuracy: 0.7349\n",
            "Epoch 51/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.7791 - accuracy: 0.7226 - val_loss: 0.8258 - val_accuracy: 0.6867\n",
            "Epoch 52/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.6920 - accuracy: 0.7660 - val_loss: 0.8404 - val_accuracy: 0.6867\n",
            "Epoch 53/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.7642 - accuracy: 0.7110 - val_loss: 0.8400 - val_accuracy: 0.6145\n",
            "Epoch 54/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.8172 - accuracy: 0.6983 - val_loss: 0.7232 - val_accuracy: 0.7470\n",
            "Epoch 55/300\n",
            "33/33 [==============================] - 15s 454ms/step - loss: 0.7183 - accuracy: 0.7193 - val_loss: 0.7548 - val_accuracy: 0.7108\n",
            "Epoch 56/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.6806 - accuracy: 0.7560 - val_loss: 0.7723 - val_accuracy: 0.6867\n",
            "Epoch 57/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.7048 - accuracy: 0.7450 - val_loss: 0.8425 - val_accuracy: 0.6747\n",
            "Epoch 58/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.7076 - accuracy: 0.7555 - val_loss: 0.7522 - val_accuracy: 0.7590\n",
            "Epoch 59/300\n",
            "33/33 [==============================] - 15s 454ms/step - loss: 0.6938 - accuracy: 0.7340 - val_loss: 0.8280 - val_accuracy: 0.6988\n",
            "Epoch 60/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.6965 - accuracy: 0.7313 - val_loss: 0.7792 - val_accuracy: 0.6747\n",
            "Epoch 61/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.6683 - accuracy: 0.7769 - val_loss: 0.7812 - val_accuracy: 0.6867\n",
            "Epoch 62/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.7015 - accuracy: 0.7410 - val_loss: 0.7945 - val_accuracy: 0.7229\n",
            "Epoch 63/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.7039 - accuracy: 0.7467 - val_loss: 0.7690 - val_accuracy: 0.6867\n",
            "Epoch 64/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.6799 - accuracy: 0.7511 - val_loss: 0.7533 - val_accuracy: 0.7349\n",
            "Epoch 65/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.7103 - accuracy: 0.7437 - val_loss: 0.8042 - val_accuracy: 0.6506\n",
            "Epoch 66/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.6909 - accuracy: 0.7739 - val_loss: 0.7696 - val_accuracy: 0.7108\n",
            "Epoch 67/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.6878 - accuracy: 0.7536 - val_loss: 0.7701 - val_accuracy: 0.6988\n",
            "Epoch 68/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.6363 - accuracy: 0.7777 - val_loss: 0.7539 - val_accuracy: 0.7108\n",
            "Epoch 69/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.6515 - accuracy: 0.7505 - val_loss: 0.7203 - val_accuracy: 0.7229\n",
            "Epoch 70/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.6240 - accuracy: 0.7834 - val_loss: 0.7375 - val_accuracy: 0.6988\n",
            "Epoch 71/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.6097 - accuracy: 0.7850 - val_loss: 0.7270 - val_accuracy: 0.7229\n",
            "Epoch 72/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.6216 - accuracy: 0.7823 - val_loss: 0.7289 - val_accuracy: 0.7349\n",
            "Epoch 73/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.6365 - accuracy: 0.7550 - val_loss: 0.7211 - val_accuracy: 0.7349\n",
            "Epoch 74/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.6291 - accuracy: 0.8034 - val_loss: 0.6965 - val_accuracy: 0.7831\n",
            "Epoch 75/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.6340 - accuracy: 0.7671 - val_loss: 0.5685 - val_accuracy: 0.7831\n",
            "Epoch 76/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.6082 - accuracy: 0.7855 - val_loss: 0.7249 - val_accuracy: 0.6627\n",
            "Epoch 77/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.6096 - accuracy: 0.7847 - val_loss: 0.6758 - val_accuracy: 0.7831\n",
            "Epoch 78/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.6291 - accuracy: 0.7753 - val_loss: 0.7298 - val_accuracy: 0.7108\n",
            "Epoch 79/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.6105 - accuracy: 0.7673 - val_loss: 0.6641 - val_accuracy: 0.7711\n",
            "Epoch 80/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.6462 - accuracy: 0.7782 - val_loss: 0.6816 - val_accuracy: 0.7952\n",
            "Epoch 81/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.6184 - accuracy: 0.7931 - val_loss: 0.7694 - val_accuracy: 0.6747\n",
            "Epoch 82/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.5451 - accuracy: 0.7973 - val_loss: 0.6786 - val_accuracy: 0.7229\n",
            "Epoch 83/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.6036 - accuracy: 0.7822 - val_loss: 0.7361 - val_accuracy: 0.7229\n",
            "Epoch 84/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.6134 - accuracy: 0.7867 - val_loss: 0.6591 - val_accuracy: 0.7470\n",
            "Epoch 85/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.5207 - accuracy: 0.8129 - val_loss: 0.6772 - val_accuracy: 0.7831\n",
            "Epoch 86/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.5708 - accuracy: 0.7736 - val_loss: 0.7272 - val_accuracy: 0.7831\n",
            "Epoch 87/300\n",
            "33/33 [==============================] - 15s 454ms/step - loss: 0.5756 - accuracy: 0.7908 - val_loss: 0.7904 - val_accuracy: 0.7229\n",
            "Epoch 88/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.5535 - accuracy: 0.8048 - val_loss: 0.6976 - val_accuracy: 0.7590\n",
            "Epoch 89/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.5806 - accuracy: 0.8118 - val_loss: 0.6410 - val_accuracy: 0.8072\n",
            "Epoch 90/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.5385 - accuracy: 0.7960 - val_loss: 0.7264 - val_accuracy: 0.7229\n",
            "Epoch 91/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.5222 - accuracy: 0.8216 - val_loss: 0.7310 - val_accuracy: 0.7470\n",
            "Epoch 92/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.5613 - accuracy: 0.7990 - val_loss: 0.6790 - val_accuracy: 0.7711\n",
            "Epoch 93/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.5910 - accuracy: 0.7916 - val_loss: 0.7129 - val_accuracy: 0.7952\n",
            "Epoch 94/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.5558 - accuracy: 0.7843 - val_loss: 0.5832 - val_accuracy: 0.8072\n",
            "Epoch 95/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.5229 - accuracy: 0.8193 - val_loss: 0.6596 - val_accuracy: 0.7831\n",
            "Epoch 96/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.5388 - accuracy: 0.8227 - val_loss: 0.6888 - val_accuracy: 0.7349\n",
            "Epoch 97/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.5161 - accuracy: 0.8176 - val_loss: 0.7498 - val_accuracy: 0.7470\n",
            "Epoch 98/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.5277 - accuracy: 0.8218 - val_loss: 0.7396 - val_accuracy: 0.7229\n",
            "Epoch 99/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.5704 - accuracy: 0.7988 - val_loss: 0.5870 - val_accuracy: 0.7711\n",
            "Epoch 100/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.5015 - accuracy: 0.8298 - val_loss: 0.7456 - val_accuracy: 0.7470\n",
            "Epoch 101/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.5161 - accuracy: 0.8215 - val_loss: 0.7639 - val_accuracy: 0.6988\n",
            "Epoch 102/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.5711 - accuracy: 0.8065 - val_loss: 0.6248 - val_accuracy: 0.7590\n",
            "Epoch 103/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.4762 - accuracy: 0.8311 - val_loss: 0.6539 - val_accuracy: 0.7711\n",
            "Epoch 104/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.5163 - accuracy: 0.8199 - val_loss: 0.7736 - val_accuracy: 0.7108\n",
            "Epoch 105/300\n",
            "33/33 [==============================] - 15s 453ms/step - loss: 0.5119 - accuracy: 0.8135 - val_loss: 0.8051 - val_accuracy: 0.7349\n",
            "Epoch 106/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.5231 - accuracy: 0.8221 - val_loss: 0.5810 - val_accuracy: 0.7831\n",
            "Epoch 107/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.5202 - accuracy: 0.8041 - val_loss: 0.6517 - val_accuracy: 0.7590\n",
            "Epoch 108/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.5072 - accuracy: 0.8396 - val_loss: 0.7280 - val_accuracy: 0.7470\n",
            "Epoch 109/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.5198 - accuracy: 0.8153 - val_loss: 0.6990 - val_accuracy: 0.7470\n",
            "Epoch 110/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.4825 - accuracy: 0.8120 - val_loss: 0.6797 - val_accuracy: 0.7229\n",
            "Epoch 111/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.4630 - accuracy: 0.8499 - val_loss: 0.6833 - val_accuracy: 0.7349\n",
            "Epoch 112/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.4951 - accuracy: 0.8204 - val_loss: 0.6518 - val_accuracy: 0.7470\n",
            "Epoch 113/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.4278 - accuracy: 0.8583 - val_loss: 0.6259 - val_accuracy: 0.7711\n",
            "Epoch 114/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.4876 - accuracy: 0.8281 - val_loss: 0.6699 - val_accuracy: 0.7590\n",
            "Epoch 115/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.4646 - accuracy: 0.8271 - val_loss: 0.6035 - val_accuracy: 0.7952\n",
            "Epoch 116/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.4985 - accuracy: 0.8231 - val_loss: 0.5799 - val_accuracy: 0.7952\n",
            "Epoch 117/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.4934 - accuracy: 0.8278 - val_loss: 0.6638 - val_accuracy: 0.7470\n",
            "Epoch 118/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.4574 - accuracy: 0.8427 - val_loss: 0.6560 - val_accuracy: 0.7590\n",
            "Epoch 119/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.5109 - accuracy: 0.8272 - val_loss: 0.6408 - val_accuracy: 0.7470\n",
            "Epoch 120/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.4345 - accuracy: 0.8546 - val_loss: 0.7317 - val_accuracy: 0.7831\n",
            "Epoch 121/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.5086 - accuracy: 0.8122 - val_loss: 0.7210 - val_accuracy: 0.7590\n",
            "Epoch 122/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.4625 - accuracy: 0.8316 - val_loss: 0.7184 - val_accuracy: 0.7590\n",
            "Epoch 123/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.4744 - accuracy: 0.8210 - val_loss: 0.6546 - val_accuracy: 0.7470\n",
            "Epoch 124/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.4967 - accuracy: 0.8245 - val_loss: 0.6755 - val_accuracy: 0.7590\n",
            "Epoch 125/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.4316 - accuracy: 0.8384 - val_loss: 0.6691 - val_accuracy: 0.7590\n",
            "Epoch 126/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.4483 - accuracy: 0.8331 - val_loss: 0.6123 - val_accuracy: 0.7831\n",
            "Epoch 127/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.4692 - accuracy: 0.8280 - val_loss: 0.6631 - val_accuracy: 0.7711\n",
            "Epoch 128/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.4620 - accuracy: 0.8397 - val_loss: 0.6129 - val_accuracy: 0.7952\n",
            "Epoch 129/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.3963 - accuracy: 0.8633 - val_loss: 0.6766 - val_accuracy: 0.7711\n",
            "Epoch 130/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.4013 - accuracy: 0.8678 - val_loss: 0.6886 - val_accuracy: 0.7470\n",
            "Epoch 131/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.3805 - accuracy: 0.8754 - val_loss: 0.5527 - val_accuracy: 0.8193\n",
            "Epoch 132/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.5037 - accuracy: 0.8200 - val_loss: 0.6200 - val_accuracy: 0.7831\n",
            "Epoch 133/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.4738 - accuracy: 0.8351 - val_loss: 0.5441 - val_accuracy: 0.8434\n",
            "Epoch 134/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.3943 - accuracy: 0.8584 - val_loss: 0.6482 - val_accuracy: 0.7952\n",
            "Epoch 135/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.4724 - accuracy: 0.8283 - val_loss: 0.7064 - val_accuracy: 0.7229\n",
            "Epoch 136/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.4506 - accuracy: 0.8431 - val_loss: 0.7736 - val_accuracy: 0.7470\n",
            "Epoch 137/300\n",
            "33/33 [==============================] - 15s 454ms/step - loss: 0.4784 - accuracy: 0.8306 - val_loss: 0.6776 - val_accuracy: 0.7470\n",
            "Epoch 138/300\n",
            "33/33 [==============================] - 15s 453ms/step - loss: 0.4268 - accuracy: 0.8507 - val_loss: 0.5266 - val_accuracy: 0.7952\n",
            "Epoch 139/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.4190 - accuracy: 0.8400 - val_loss: 0.6551 - val_accuracy: 0.7831\n",
            "Epoch 140/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.3821 - accuracy: 0.8748 - val_loss: 0.6010 - val_accuracy: 0.7952\n",
            "Epoch 141/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.4228 - accuracy: 0.8613 - val_loss: 0.5554 - val_accuracy: 0.7831\n",
            "Epoch 142/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.3817 - accuracy: 0.8546 - val_loss: 0.7181 - val_accuracy: 0.7590\n",
            "Epoch 143/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.3802 - accuracy: 0.8686 - val_loss: 0.5340 - val_accuracy: 0.7831\n",
            "Epoch 144/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.3819 - accuracy: 0.8681 - val_loss: 0.6834 - val_accuracy: 0.7349\n",
            "Epoch 145/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.4564 - accuracy: 0.8660 - val_loss: 0.6656 - val_accuracy: 0.7831\n",
            "Epoch 146/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.3997 - accuracy: 0.8579 - val_loss: 0.6468 - val_accuracy: 0.7711\n",
            "Epoch 147/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.4696 - accuracy: 0.8138 - val_loss: 0.5642 - val_accuracy: 0.8313\n",
            "Epoch 148/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.4116 - accuracy: 0.8588 - val_loss: 0.6186 - val_accuracy: 0.7590\n",
            "Epoch 149/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.3925 - accuracy: 0.8724 - val_loss: 0.6217 - val_accuracy: 0.8313\n",
            "Epoch 150/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.4200 - accuracy: 0.8577 - val_loss: 0.6004 - val_accuracy: 0.8313\n",
            "Epoch 151/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.3809 - accuracy: 0.8662 - val_loss: 0.6095 - val_accuracy: 0.8193\n",
            "Epoch 152/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3772 - accuracy: 0.8620 - val_loss: 0.6047 - val_accuracy: 0.7711\n",
            "Epoch 153/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.4072 - accuracy: 0.8575 - val_loss: 0.6151 - val_accuracy: 0.7952\n",
            "Epoch 154/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.4240 - accuracy: 0.8530 - val_loss: 0.5691 - val_accuracy: 0.7952\n",
            "Epoch 155/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.4518 - accuracy: 0.8454 - val_loss: 0.7265 - val_accuracy: 0.7952\n",
            "Epoch 156/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.4315 - accuracy: 0.8432 - val_loss: 0.7414 - val_accuracy: 0.7470\n",
            "Epoch 157/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.3864 - accuracy: 0.8550 - val_loss: 0.7406 - val_accuracy: 0.7952\n",
            "Epoch 158/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.3882 - accuracy: 0.8570 - val_loss: 0.5517 - val_accuracy: 0.7952\n",
            "Epoch 159/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.3892 - accuracy: 0.8596 - val_loss: 0.6320 - val_accuracy: 0.7952\n",
            "Epoch 160/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.4055 - accuracy: 0.8375 - val_loss: 0.7381 - val_accuracy: 0.7470\n",
            "Epoch 161/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.3389 - accuracy: 0.8798 - val_loss: 0.6113 - val_accuracy: 0.7831\n",
            "Epoch 162/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.3473 - accuracy: 0.8690 - val_loss: 0.7458 - val_accuracy: 0.7711\n",
            "Epoch 163/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.3518 - accuracy: 0.8593 - val_loss: 0.6578 - val_accuracy: 0.7711\n",
            "Epoch 164/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.3909 - accuracy: 0.8665 - val_loss: 0.6314 - val_accuracy: 0.7590\n",
            "Epoch 165/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.3912 - accuracy: 0.8657 - val_loss: 0.6638 - val_accuracy: 0.7711\n",
            "Epoch 166/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.3489 - accuracy: 0.8706 - val_loss: 0.7570 - val_accuracy: 0.7590\n",
            "Epoch 167/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.4121 - accuracy: 0.8614 - val_loss: 0.5670 - val_accuracy: 0.7711\n",
            "Epoch 168/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.3701 - accuracy: 0.8819 - val_loss: 0.6845 - val_accuracy: 0.7952\n",
            "Epoch 169/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.3314 - accuracy: 0.8910 - val_loss: 0.5789 - val_accuracy: 0.7952\n",
            "Epoch 170/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.3869 - accuracy: 0.8517 - val_loss: 0.6627 - val_accuracy: 0.7108\n",
            "Epoch 171/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3271 - accuracy: 0.8953 - val_loss: 0.5861 - val_accuracy: 0.8434\n",
            "Epoch 172/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.3833 - accuracy: 0.8551 - val_loss: 0.7927 - val_accuracy: 0.8193\n",
            "Epoch 173/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.3772 - accuracy: 0.8660 - val_loss: 0.5791 - val_accuracy: 0.7711\n",
            "Epoch 174/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.3951 - accuracy: 0.8401 - val_loss: 0.7487 - val_accuracy: 0.7590\n",
            "Epoch 175/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3644 - accuracy: 0.8700 - val_loss: 0.7419 - val_accuracy: 0.8193\n",
            "Epoch 176/300\n",
            "33/33 [==============================] - 15s 472ms/step - loss: 0.3578 - accuracy: 0.8681 - val_loss: 0.7014 - val_accuracy: 0.7831\n",
            "Epoch 177/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.3192 - accuracy: 0.8949 - val_loss: 0.6683 - val_accuracy: 0.7711\n",
            "Epoch 178/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.3361 - accuracy: 0.8923 - val_loss: 0.6329 - val_accuracy: 0.7952\n",
            "Epoch 179/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.3506 - accuracy: 0.8643 - val_loss: 0.7532 - val_accuracy: 0.7470\n",
            "Epoch 180/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.3902 - accuracy: 0.8582 - val_loss: 0.5599 - val_accuracy: 0.7952\n",
            "Epoch 181/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3689 - accuracy: 0.8717 - val_loss: 0.7125 - val_accuracy: 0.7831\n",
            "Epoch 182/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3060 - accuracy: 0.8942 - val_loss: 0.4494 - val_accuracy: 0.8313\n",
            "Epoch 183/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3422 - accuracy: 0.8822 - val_loss: 0.6596 - val_accuracy: 0.8072\n",
            "Epoch 184/300\n",
            "33/33 [==============================] - 16s 466ms/step - loss: 0.3270 - accuracy: 0.8793 - val_loss: 0.7774 - val_accuracy: 0.7590\n",
            "Epoch 185/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3199 - accuracy: 0.8847 - val_loss: 0.6763 - val_accuracy: 0.7711\n",
            "Epoch 186/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.3131 - accuracy: 0.8859 - val_loss: 0.6710 - val_accuracy: 0.7831\n",
            "Epoch 187/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.3510 - accuracy: 0.8752 - val_loss: 0.6886 - val_accuracy: 0.8193\n",
            "Epoch 188/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.3217 - accuracy: 0.8881 - val_loss: 0.8648 - val_accuracy: 0.6988\n",
            "Epoch 189/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.3123 - accuracy: 0.9056 - val_loss: 0.6616 - val_accuracy: 0.7831\n",
            "Epoch 190/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.3130 - accuracy: 0.8872 - val_loss: 0.7536 - val_accuracy: 0.7831\n",
            "Epoch 191/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.3605 - accuracy: 0.8860 - val_loss: 0.6628 - val_accuracy: 0.7711\n",
            "Epoch 192/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3841 - accuracy: 0.8627 - val_loss: 0.6358 - val_accuracy: 0.7952\n",
            "Epoch 193/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.3486 - accuracy: 0.8759 - val_loss: 0.6589 - val_accuracy: 0.8072\n",
            "Epoch 194/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.3148 - accuracy: 0.8798 - val_loss: 0.6785 - val_accuracy: 0.7831\n",
            "Epoch 195/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.3326 - accuracy: 0.8975 - val_loss: 0.6194 - val_accuracy: 0.7952\n",
            "Epoch 196/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.3952 - accuracy: 0.8625 - val_loss: 0.6244 - val_accuracy: 0.8072\n",
            "Epoch 197/300\n",
            "33/33 [==============================] - 16s 476ms/step - loss: 0.3810 - accuracy: 0.8452 - val_loss: 0.7133 - val_accuracy: 0.7711\n",
            "Epoch 198/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.3049 - accuracy: 0.8868 - val_loss: 0.6099 - val_accuracy: 0.8072\n",
            "Epoch 199/300\n",
            "33/33 [==============================] - 15s 469ms/step - loss: 0.3062 - accuracy: 0.8958 - val_loss: 0.5245 - val_accuracy: 0.8193\n",
            "Epoch 200/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.3238 - accuracy: 0.8893 - val_loss: 0.8489 - val_accuracy: 0.7470\n",
            "Epoch 201/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.2835 - accuracy: 0.9054 - val_loss: 0.7108 - val_accuracy: 0.7952\n",
            "Epoch 202/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.3050 - accuracy: 0.8905 - val_loss: 0.7424 - val_accuracy: 0.7952\n",
            "Epoch 203/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3363 - accuracy: 0.8895 - val_loss: 0.6639 - val_accuracy: 0.7711\n",
            "Epoch 204/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.3740 - accuracy: 0.8704 - val_loss: 0.5965 - val_accuracy: 0.8193\n",
            "Epoch 205/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.3346 - accuracy: 0.8843 - val_loss: 0.5262 - val_accuracy: 0.8193\n",
            "Epoch 206/300\n",
            "33/33 [==============================] - 16s 473ms/step - loss: 0.2863 - accuracy: 0.8954 - val_loss: 0.6343 - val_accuracy: 0.8072\n",
            "Epoch 207/300\n",
            "33/33 [==============================] - 15s 469ms/step - loss: 0.3150 - accuracy: 0.8910 - val_loss: 0.5317 - val_accuracy: 0.8434\n",
            "Epoch 208/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.3306 - accuracy: 0.8731 - val_loss: 0.6277 - val_accuracy: 0.7952\n",
            "Epoch 209/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.2566 - accuracy: 0.9096 - val_loss: 0.7373 - val_accuracy: 0.7952\n",
            "Epoch 210/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.3313 - accuracy: 0.8801 - val_loss: 0.7923 - val_accuracy: 0.7831\n",
            "Epoch 211/300\n",
            "33/33 [==============================] - 16s 473ms/step - loss: 0.3320 - accuracy: 0.8885 - val_loss: 0.5895 - val_accuracy: 0.8313\n",
            "Epoch 212/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.3382 - accuracy: 0.8733 - val_loss: 0.6464 - val_accuracy: 0.8072\n",
            "Epoch 213/300\n",
            "33/33 [==============================] - 16s 469ms/step - loss: 0.3031 - accuracy: 0.8937 - val_loss: 0.5992 - val_accuracy: 0.7470\n",
            "Epoch 214/300\n",
            "33/33 [==============================] - 15s 473ms/step - loss: 0.3394 - accuracy: 0.8862 - val_loss: 0.6065 - val_accuracy: 0.7952\n",
            "Epoch 215/300\n",
            "33/33 [==============================] - 16s 469ms/step - loss: 0.2613 - accuracy: 0.9091 - val_loss: 0.8247 - val_accuracy: 0.7590\n",
            "Epoch 216/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.3342 - accuracy: 0.8756 - val_loss: 0.7571 - val_accuracy: 0.7470\n",
            "Epoch 217/300\n",
            "33/33 [==============================] - 16s 478ms/step - loss: 0.3720 - accuracy: 0.8756 - val_loss: 0.7426 - val_accuracy: 0.7711\n",
            "Epoch 218/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.2734 - accuracy: 0.9089 - val_loss: 0.8598 - val_accuracy: 0.7590\n",
            "Epoch 219/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.3112 - accuracy: 0.8935 - val_loss: 0.8647 - val_accuracy: 0.7349\n",
            "Epoch 220/300\n",
            "33/33 [==============================] - 16s 473ms/step - loss: 0.2868 - accuracy: 0.9007 - val_loss: 0.6679 - val_accuracy: 0.7590\n",
            "Epoch 221/300\n",
            "33/33 [==============================] - 16s 470ms/step - loss: 0.3391 - accuracy: 0.8823 - val_loss: 0.7130 - val_accuracy: 0.7831\n",
            "Epoch 222/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.3335 - accuracy: 0.8654 - val_loss: 0.6478 - val_accuracy: 0.7711\n",
            "Epoch 223/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.3114 - accuracy: 0.8918 - val_loss: 0.5741 - val_accuracy: 0.7952\n",
            "Epoch 224/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.3235 - accuracy: 0.8832 - val_loss: 0.5598 - val_accuracy: 0.8434\n",
            "Epoch 225/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.2890 - accuracy: 0.8874 - val_loss: 0.6317 - val_accuracy: 0.7952\n",
            "Epoch 226/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.2914 - accuracy: 0.8981 - val_loss: 0.6064 - val_accuracy: 0.8072\n",
            "Epoch 227/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.3147 - accuracy: 0.8954 - val_loss: 0.5299 - val_accuracy: 0.8313\n",
            "Epoch 228/300\n",
            "33/33 [==============================] - 16s 468ms/step - loss: 0.2971 - accuracy: 0.8816 - val_loss: 0.6924 - val_accuracy: 0.7711\n",
            "Epoch 229/300\n",
            "33/33 [==============================] - 16s 473ms/step - loss: 0.3106 - accuracy: 0.8822 - val_loss: 0.6072 - val_accuracy: 0.7952\n",
            "Epoch 230/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.2830 - accuracy: 0.8975 - val_loss: 0.7023 - val_accuracy: 0.7711\n",
            "Epoch 231/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.2919 - accuracy: 0.8952 - val_loss: 0.6624 - val_accuracy: 0.8072\n",
            "Epoch 232/300\n",
            "33/33 [==============================] - 16s 469ms/step - loss: 0.2601 - accuracy: 0.9137 - val_loss: 0.6370 - val_accuracy: 0.7952\n",
            "Epoch 233/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.2739 - accuracy: 0.9028 - val_loss: 0.5732 - val_accuracy: 0.8072\n",
            "Epoch 234/300\n",
            "33/33 [==============================] - 16s 473ms/step - loss: 0.3470 - accuracy: 0.8727 - val_loss: 0.6089 - val_accuracy: 0.8193\n",
            "Epoch 235/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.3191 - accuracy: 0.8897 - val_loss: 0.5420 - val_accuracy: 0.8072\n",
            "Epoch 236/300\n",
            "33/33 [==============================] - 16s 476ms/step - loss: 0.3110 - accuracy: 0.8762 - val_loss: 0.6889 - val_accuracy: 0.8313\n",
            "Epoch 237/300\n",
            "33/33 [==============================] - 16s 478ms/step - loss: 0.3243 - accuracy: 0.8856 - val_loss: 0.5287 - val_accuracy: 0.7952\n",
            "Epoch 238/300\n",
            "33/33 [==============================] - 16s 469ms/step - loss: 0.3216 - accuracy: 0.8959 - val_loss: 0.6060 - val_accuracy: 0.8313\n",
            "Epoch 239/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.2664 - accuracy: 0.9027 - val_loss: 0.6748 - val_accuracy: 0.8072\n",
            "Epoch 240/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2650 - accuracy: 0.9013 - val_loss: 0.6277 - val_accuracy: 0.7831\n",
            "Epoch 241/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.2601 - accuracy: 0.9162 - val_loss: 0.6706 - val_accuracy: 0.7831\n",
            "Epoch 242/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.3127 - accuracy: 0.8863 - val_loss: 0.7420 - val_accuracy: 0.8072\n",
            "Epoch 243/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.3063 - accuracy: 0.8973 - val_loss: 0.5396 - val_accuracy: 0.8193\n",
            "Epoch 244/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2805 - accuracy: 0.9021 - val_loss: 0.5160 - val_accuracy: 0.8313\n",
            "Epoch 245/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.3023 - accuracy: 0.8781 - val_loss: 0.5938 - val_accuracy: 0.8434\n",
            "Epoch 246/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2929 - accuracy: 0.8894 - val_loss: 0.8028 - val_accuracy: 0.7470\n",
            "Epoch 247/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2917 - accuracy: 0.9091 - val_loss: 0.5364 - val_accuracy: 0.8434\n",
            "Epoch 248/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3021 - accuracy: 0.8886 - val_loss: 0.6055 - val_accuracy: 0.8072\n",
            "Epoch 249/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2723 - accuracy: 0.9021 - val_loss: 0.5935 - val_accuracy: 0.7831\n",
            "Epoch 250/300\n",
            "33/33 [==============================] - 15s 473ms/step - loss: 0.3037 - accuracy: 0.8966 - val_loss: 0.6079 - val_accuracy: 0.8072\n",
            "Epoch 251/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2598 - accuracy: 0.9063 - val_loss: 0.6840 - val_accuracy: 0.7831\n",
            "Epoch 252/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.2794 - accuracy: 0.8961 - val_loss: 0.6861 - val_accuracy: 0.8193\n",
            "Epoch 253/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2872 - accuracy: 0.8917 - val_loss: 0.5729 - val_accuracy: 0.7952\n",
            "Epoch 254/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2885 - accuracy: 0.8905 - val_loss: 0.6248 - val_accuracy: 0.7952\n",
            "Epoch 255/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.3145 - accuracy: 0.8882 - val_loss: 0.5402 - val_accuracy: 0.8434\n",
            "Epoch 256/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2915 - accuracy: 0.8905 - val_loss: 0.4943 - val_accuracy: 0.8795\n",
            "Epoch 257/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.2394 - accuracy: 0.9218 - val_loss: 0.5947 - val_accuracy: 0.8072\n",
            "Epoch 258/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2472 - accuracy: 0.9215 - val_loss: 0.8091 - val_accuracy: 0.7831\n",
            "Epoch 259/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2799 - accuracy: 0.8965 - val_loss: 0.6828 - val_accuracy: 0.7711\n",
            "Epoch 260/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.3128 - accuracy: 0.9040 - val_loss: 0.6754 - val_accuracy: 0.8554\n",
            "Epoch 261/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2601 - accuracy: 0.9014 - val_loss: 0.7404 - val_accuracy: 0.7711\n",
            "Epoch 262/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2888 - accuracy: 0.9040 - val_loss: 0.7451 - val_accuracy: 0.8193\n",
            "Epoch 263/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2579 - accuracy: 0.9082 - val_loss: 0.5603 - val_accuracy: 0.8313\n",
            "Epoch 264/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2517 - accuracy: 0.9103 - val_loss: 0.6955 - val_accuracy: 0.8193\n",
            "Epoch 265/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2765 - accuracy: 0.9143 - val_loss: 0.5339 - val_accuracy: 0.8072\n",
            "Epoch 266/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2498 - accuracy: 0.9094 - val_loss: 0.5564 - val_accuracy: 0.8434\n",
            "Epoch 267/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2372 - accuracy: 0.9017 - val_loss: 0.6768 - val_accuracy: 0.7711\n",
            "Epoch 268/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2631 - accuracy: 0.9065 - val_loss: 0.6464 - val_accuracy: 0.8193\n",
            "Epoch 269/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3163 - accuracy: 0.8886 - val_loss: 0.5778 - val_accuracy: 0.8434\n",
            "Epoch 270/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3069 - accuracy: 0.8864 - val_loss: 0.7119 - val_accuracy: 0.7952\n",
            "Epoch 271/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2532 - accuracy: 0.9211 - val_loss: 0.7524 - val_accuracy: 0.7590\n",
            "Epoch 272/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2555 - accuracy: 0.8988 - val_loss: 0.7431 - val_accuracy: 0.7952\n",
            "Epoch 273/300\n",
            "33/33 [==============================] - 15s 469ms/step - loss: 0.2679 - accuracy: 0.9100 - val_loss: 0.7656 - val_accuracy: 0.7711\n",
            "Epoch 274/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2676 - accuracy: 0.9113 - val_loss: 0.5433 - val_accuracy: 0.8072\n",
            "Epoch 275/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.2654 - accuracy: 0.9062 - val_loss: 0.6571 - val_accuracy: 0.7590\n",
            "Epoch 276/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2123 - accuracy: 0.9218 - val_loss: 0.7566 - val_accuracy: 0.8193\n",
            "Epoch 277/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.2637 - accuracy: 0.9080 - val_loss: 0.5117 - val_accuracy: 0.8554\n",
            "Epoch 278/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.2537 - accuracy: 0.9204 - val_loss: 0.8595 - val_accuracy: 0.7229\n",
            "Epoch 279/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.3049 - accuracy: 0.8990 - val_loss: 0.5799 - val_accuracy: 0.8193\n",
            "Epoch 280/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2835 - accuracy: 0.9002 - val_loss: 0.7109 - val_accuracy: 0.8313\n",
            "Epoch 281/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.3163 - accuracy: 0.8980 - val_loss: 0.5570 - val_accuracy: 0.8313\n",
            "Epoch 282/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2792 - accuracy: 0.9047 - val_loss: 0.7241 - val_accuracy: 0.7831\n",
            "Epoch 283/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.3032 - accuracy: 0.8930 - val_loss: 0.5607 - val_accuracy: 0.7831\n",
            "Epoch 284/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.3090 - accuracy: 0.8927 - val_loss: 0.8113 - val_accuracy: 0.7470\n",
            "Epoch 285/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2530 - accuracy: 0.9045 - val_loss: 0.7116 - val_accuracy: 0.8434\n",
            "Epoch 286/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2698 - accuracy: 0.9029 - val_loss: 0.6889 - val_accuracy: 0.8554\n",
            "Epoch 287/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.3015 - accuracy: 0.8800 - val_loss: 0.8098 - val_accuracy: 0.8193\n",
            "Epoch 288/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2742 - accuracy: 0.8966 - val_loss: 0.8392 - val_accuracy: 0.7470\n",
            "Epoch 289/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2312 - accuracy: 0.9199 - val_loss: 0.5678 - val_accuracy: 0.8313\n",
            "Epoch 290/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2577 - accuracy: 0.9135 - val_loss: 0.5968 - val_accuracy: 0.8193\n",
            "Epoch 291/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2200 - accuracy: 0.9153 - val_loss: 0.6909 - val_accuracy: 0.7470\n",
            "Epoch 292/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2800 - accuracy: 0.8865 - val_loss: 0.7946 - val_accuracy: 0.7590\n",
            "Epoch 293/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2608 - accuracy: 0.9110 - val_loss: 0.6267 - val_accuracy: 0.7952\n",
            "Epoch 294/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.2410 - accuracy: 0.9100 - val_loss: 0.7729 - val_accuracy: 0.6988\n",
            "Epoch 295/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2459 - accuracy: 0.9008 - val_loss: 0.5931 - val_accuracy: 0.8313\n",
            "Epoch 296/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2283 - accuracy: 0.9228 - val_loss: 0.6641 - val_accuracy: 0.8072\n",
            "Epoch 297/300\n",
            "33/33 [==============================] - 15s 469ms/step - loss: 0.2691 - accuracy: 0.9157 - val_loss: 0.7484 - val_accuracy: 0.8072\n",
            "Epoch 298/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2776 - accuracy: 0.9179 - val_loss: 0.7610 - val_accuracy: 0.7831\n",
            "Epoch 299/300\n",
            "33/33 [==============================] - 16s 468ms/step - loss: 0.2160 - accuracy: 0.9200 - val_loss: 0.6537 - val_accuracy: 0.8313\n",
            "Epoch 300/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2266 - accuracy: 0.9216 - val_loss: 0.6589 - val_accuracy: 0.8313\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gi4lchPvBYKg",
        "outputId": "557edae5-f8be-47bc-eb10-8f6847f7966b"
      },
      "source": [
        "final_loss, final_accuracy = model.evaluate(valid_generator)\r\n",
        "print(\"Final loss: {:.2f}\".format(final_loss))\r\n",
        "print(\"Final accuracy: {:.2f}%\".format(final_accuracy * 100))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "3/3 [==============================] - 1s 380ms/step - loss: 0.5991 - accuracy: 0.7952\n",
            "Final loss: 0.60\n",
            "Final accuracy: 79.52%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NLW6gyGUaqM1",
        "outputId": "0320a4d0-8ecd-4b76-c103-49e72e275109"
      },
      "source": [
        "hist= model.fit(train_generator, validation_data = valid_generator, epochs = 300)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/300\n",
            "33/33 [==============================] - 16s 477ms/step - loss: 0.2361 - accuracy: 0.9148 - val_loss: 0.5991 - val_accuracy: 0.7952\n",
            "Epoch 2/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2481 - accuracy: 0.9053 - val_loss: 0.7662 - val_accuracy: 0.7470\n",
            "Epoch 3/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2838 - accuracy: 0.8890 - val_loss: 0.8643 - val_accuracy: 0.7952\n",
            "Epoch 4/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2541 - accuracy: 0.9062 - val_loss: 0.6146 - val_accuracy: 0.8193\n",
            "Epoch 5/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2338 - accuracy: 0.9234 - val_loss: 0.5626 - val_accuracy: 0.8434\n",
            "Epoch 6/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2751 - accuracy: 0.9053 - val_loss: 0.6870 - val_accuracy: 0.7831\n",
            "Epoch 7/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2584 - accuracy: 0.9100 - val_loss: 0.6547 - val_accuracy: 0.8313\n",
            "Epoch 8/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2509 - accuracy: 0.9091 - val_loss: 0.6355 - val_accuracy: 0.8193\n",
            "Epoch 9/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2784 - accuracy: 0.8967 - val_loss: 0.6006 - val_accuracy: 0.8554\n",
            "Epoch 10/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2627 - accuracy: 0.9005 - val_loss: 0.7654 - val_accuracy: 0.7952\n",
            "Epoch 11/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2617 - accuracy: 0.9014 - val_loss: 0.5756 - val_accuracy: 0.8072\n",
            "Epoch 12/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.2437 - accuracy: 0.9024 - val_loss: 0.5633 - val_accuracy: 0.8072\n",
            "Epoch 13/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2682 - accuracy: 0.8986 - val_loss: 0.7001 - val_accuracy: 0.8313\n",
            "Epoch 14/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.2392 - accuracy: 0.9129 - val_loss: 0.7599 - val_accuracy: 0.8072\n",
            "Epoch 15/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2455 - accuracy: 0.9091 - val_loss: 0.6198 - val_accuracy: 0.8072\n",
            "Epoch 16/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2534 - accuracy: 0.9110 - val_loss: 0.8197 - val_accuracy: 0.7590\n",
            "Epoch 17/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2390 - accuracy: 0.9110 - val_loss: 0.5106 - val_accuracy: 0.8313\n",
            "Epoch 18/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2492 - accuracy: 0.9129 - val_loss: 0.7413 - val_accuracy: 0.7470\n",
            "Epoch 19/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2303 - accuracy: 0.9139 - val_loss: 0.8410 - val_accuracy: 0.7711\n",
            "Epoch 20/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2588 - accuracy: 0.9014 - val_loss: 0.7268 - val_accuracy: 0.8434\n",
            "Epoch 21/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2349 - accuracy: 0.9177 - val_loss: 0.6485 - val_accuracy: 0.8554\n",
            "Epoch 22/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.2299 - accuracy: 0.9120 - val_loss: 0.7774 - val_accuracy: 0.7711\n",
            "Epoch 23/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2359 - accuracy: 0.9234 - val_loss: 0.6368 - val_accuracy: 0.8072\n",
            "Epoch 24/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.2626 - accuracy: 0.9014 - val_loss: 0.7819 - val_accuracy: 0.7831\n",
            "Epoch 25/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2437 - accuracy: 0.9167 - val_loss: 0.7360 - val_accuracy: 0.8072\n",
            "Epoch 26/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2452 - accuracy: 0.9110 - val_loss: 0.7949 - val_accuracy: 0.8072\n",
            "Epoch 27/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2336 - accuracy: 0.9100 - val_loss: 0.7872 - val_accuracy: 0.8072\n",
            "Epoch 28/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.2698 - accuracy: 0.9187 - val_loss: 0.6587 - val_accuracy: 0.7952\n",
            "Epoch 29/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2608 - accuracy: 0.9072 - val_loss: 0.8031 - val_accuracy: 0.7831\n",
            "Epoch 30/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2630 - accuracy: 0.9072 - val_loss: 0.7791 - val_accuracy: 0.7711\n",
            "Epoch 31/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2114 - accuracy: 0.9330 - val_loss: 0.8778 - val_accuracy: 0.7831\n",
            "Epoch 32/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2675 - accuracy: 0.9110 - val_loss: 0.6603 - val_accuracy: 0.7831\n",
            "Epoch 33/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.2215 - accuracy: 0.9139 - val_loss: 0.7567 - val_accuracy: 0.7711\n",
            "Epoch 34/300\n",
            "33/33 [==============================] - 15s 454ms/step - loss: 0.2430 - accuracy: 0.9177 - val_loss: 0.6078 - val_accuracy: 0.7711\n",
            "Epoch 35/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1883 - accuracy: 0.9349 - val_loss: 0.7638 - val_accuracy: 0.7831\n",
            "Epoch 36/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2476 - accuracy: 0.9187 - val_loss: 0.9598 - val_accuracy: 0.7470\n",
            "Epoch 37/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2263 - accuracy: 0.9273 - val_loss: 0.7336 - val_accuracy: 0.8434\n",
            "Epoch 38/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.2185 - accuracy: 0.9225 - val_loss: 0.7427 - val_accuracy: 0.7952\n",
            "Epoch 39/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2400 - accuracy: 0.9062 - val_loss: 0.6268 - val_accuracy: 0.8193\n",
            "Epoch 40/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2391 - accuracy: 0.9148 - val_loss: 0.5691 - val_accuracy: 0.8313\n",
            "Epoch 41/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2514 - accuracy: 0.9043 - val_loss: 0.8315 - val_accuracy: 0.7711\n",
            "Epoch 42/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2598 - accuracy: 0.9129 - val_loss: 0.8260 - val_accuracy: 0.7831\n",
            "Epoch 43/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2039 - accuracy: 0.9196 - val_loss: 0.6659 - val_accuracy: 0.8313\n",
            "Epoch 44/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.2458 - accuracy: 0.9158 - val_loss: 0.8324 - val_accuracy: 0.7831\n",
            "Epoch 45/300\n",
            "33/33 [==============================] - 15s 473ms/step - loss: 0.2779 - accuracy: 0.9024 - val_loss: 0.7623 - val_accuracy: 0.8313\n",
            "Epoch 46/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2444 - accuracy: 0.9225 - val_loss: 0.6531 - val_accuracy: 0.7831\n",
            "Epoch 47/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2020 - accuracy: 0.9282 - val_loss: 0.6629 - val_accuracy: 0.8072\n",
            "Epoch 48/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2124 - accuracy: 0.9311 - val_loss: 0.8465 - val_accuracy: 0.8193\n",
            "Epoch 49/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2292 - accuracy: 0.9234 - val_loss: 0.5827 - val_accuracy: 0.8072\n",
            "Epoch 50/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2159 - accuracy: 0.9273 - val_loss: 0.7517 - val_accuracy: 0.8072\n",
            "Epoch 51/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2290 - accuracy: 0.9129 - val_loss: 0.6630 - val_accuracy: 0.8313\n",
            "Epoch 52/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2546 - accuracy: 0.9167 - val_loss: 0.7318 - val_accuracy: 0.8193\n",
            "Epoch 53/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2226 - accuracy: 0.9263 - val_loss: 0.9085 - val_accuracy: 0.7711\n",
            "Epoch 54/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2031 - accuracy: 0.9282 - val_loss: 0.9466 - val_accuracy: 0.7349\n",
            "Epoch 55/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2209 - accuracy: 0.9206 - val_loss: 0.6922 - val_accuracy: 0.8193\n",
            "Epoch 56/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.2239 - accuracy: 0.9100 - val_loss: 0.6908 - val_accuracy: 0.8072\n",
            "Epoch 57/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2021 - accuracy: 0.9301 - val_loss: 0.7041 - val_accuracy: 0.8434\n",
            "Epoch 58/300\n",
            "33/33 [==============================] - 16s 471ms/step - loss: 0.2041 - accuracy: 0.9263 - val_loss: 0.7166 - val_accuracy: 0.8072\n",
            "Epoch 59/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2476 - accuracy: 0.9187 - val_loss: 0.8079 - val_accuracy: 0.7831\n",
            "Epoch 60/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2373 - accuracy: 0.9177 - val_loss: 0.6065 - val_accuracy: 0.8193\n",
            "Epoch 61/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2247 - accuracy: 0.9187 - val_loss: 0.7379 - val_accuracy: 0.8193\n",
            "Epoch 62/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2146 - accuracy: 0.9244 - val_loss: 0.6562 - val_accuracy: 0.8072\n",
            "Epoch 63/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2097 - accuracy: 0.9177 - val_loss: 0.7328 - val_accuracy: 0.7952\n",
            "Epoch 64/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2005 - accuracy: 0.9244 - val_loss: 0.9122 - val_accuracy: 0.7831\n",
            "Epoch 65/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1964 - accuracy: 0.9321 - val_loss: 0.4351 - val_accuracy: 0.8554\n",
            "Epoch 66/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2244 - accuracy: 0.9158 - val_loss: 0.6811 - val_accuracy: 0.7711\n",
            "Epoch 67/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2014 - accuracy: 0.9301 - val_loss: 0.6598 - val_accuracy: 0.7831\n",
            "Epoch 68/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2113 - accuracy: 0.9215 - val_loss: 0.7452 - val_accuracy: 0.7952\n",
            "Epoch 69/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1978 - accuracy: 0.9301 - val_loss: 0.9636 - val_accuracy: 0.7952\n",
            "Epoch 70/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2120 - accuracy: 0.9301 - val_loss: 0.6858 - val_accuracy: 0.7470\n",
            "Epoch 71/300\n",
            "33/33 [==============================] - 15s 455ms/step - loss: 0.2073 - accuracy: 0.9206 - val_loss: 0.7178 - val_accuracy: 0.8313\n",
            "Epoch 72/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2554 - accuracy: 0.9148 - val_loss: 0.8596 - val_accuracy: 0.8072\n",
            "Epoch 73/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2241 - accuracy: 0.9254 - val_loss: 0.7378 - val_accuracy: 0.7711\n",
            "Epoch 74/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2158 - accuracy: 0.9263 - val_loss: 0.7027 - val_accuracy: 0.8072\n",
            "Epoch 75/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2133 - accuracy: 0.9244 - val_loss: 0.7891 - val_accuracy: 0.7831\n",
            "Epoch 76/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2215 - accuracy: 0.9196 - val_loss: 0.7501 - val_accuracy: 0.8193\n",
            "Epoch 77/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2040 - accuracy: 0.9282 - val_loss: 0.5863 - val_accuracy: 0.8554\n",
            "Epoch 78/300\n",
            "33/33 [==============================] - 15s 467ms/step - loss: 0.2309 - accuracy: 0.9254 - val_loss: 0.6996 - val_accuracy: 0.8072\n",
            "Epoch 79/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2372 - accuracy: 0.9273 - val_loss: 0.7764 - val_accuracy: 0.8193\n",
            "Epoch 80/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.2377 - accuracy: 0.9158 - val_loss: 0.6149 - val_accuracy: 0.8313\n",
            "Epoch 81/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2278 - accuracy: 0.9062 - val_loss: 0.8843 - val_accuracy: 0.7590\n",
            "Epoch 82/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.1872 - accuracy: 0.9368 - val_loss: 0.6084 - val_accuracy: 0.8313\n",
            "Epoch 83/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1880 - accuracy: 0.9225 - val_loss: 0.7581 - val_accuracy: 0.7952\n",
            "Epoch 84/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2120 - accuracy: 0.9148 - val_loss: 0.7648 - val_accuracy: 0.8072\n",
            "Epoch 85/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1938 - accuracy: 0.9349 - val_loss: 0.6783 - val_accuracy: 0.8313\n",
            "Epoch 86/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2283 - accuracy: 0.9215 - val_loss: 0.8670 - val_accuracy: 0.8072\n",
            "Epoch 87/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2121 - accuracy: 0.9196 - val_loss: 0.6055 - val_accuracy: 0.8313\n",
            "Epoch 88/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2155 - accuracy: 0.9225 - val_loss: 0.7293 - val_accuracy: 0.8193\n",
            "Epoch 89/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2270 - accuracy: 0.9282 - val_loss: 0.5685 - val_accuracy: 0.8434\n",
            "Epoch 90/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2535 - accuracy: 0.9081 - val_loss: 0.7671 - val_accuracy: 0.8072\n",
            "Epoch 91/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2282 - accuracy: 0.9167 - val_loss: 0.5545 - val_accuracy: 0.8072\n",
            "Epoch 92/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2406 - accuracy: 0.9062 - val_loss: 0.5313 - val_accuracy: 0.8554\n",
            "Epoch 93/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2271 - accuracy: 0.9263 - val_loss: 0.6683 - val_accuracy: 0.8434\n",
            "Epoch 94/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2270 - accuracy: 0.9148 - val_loss: 0.7345 - val_accuracy: 0.8193\n",
            "Epoch 95/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2341 - accuracy: 0.9215 - val_loss: 0.7893 - val_accuracy: 0.8072\n",
            "Epoch 96/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2200 - accuracy: 0.9282 - val_loss: 0.8163 - val_accuracy: 0.7952\n",
            "Epoch 97/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2319 - accuracy: 0.9129 - val_loss: 0.8169 - val_accuracy: 0.7952\n",
            "Epoch 98/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2522 - accuracy: 0.8986 - val_loss: 0.7280 - val_accuracy: 0.8554\n",
            "Epoch 99/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.1879 - accuracy: 0.9321 - val_loss: 0.9282 - val_accuracy: 0.7590\n",
            "Epoch 100/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.2273 - accuracy: 0.9206 - val_loss: 0.6706 - val_accuracy: 0.8313\n",
            "Epoch 101/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2157 - accuracy: 0.9196 - val_loss: 0.6030 - val_accuracy: 0.8313\n",
            "Epoch 102/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2232 - accuracy: 0.9273 - val_loss: 0.7171 - val_accuracy: 0.8193\n",
            "Epoch 103/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2696 - accuracy: 0.9062 - val_loss: 0.7218 - val_accuracy: 0.7952\n",
            "Epoch 104/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2121 - accuracy: 0.9263 - val_loss: 0.6712 - val_accuracy: 0.8313\n",
            "Epoch 105/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2238 - accuracy: 0.9187 - val_loss: 0.9473 - val_accuracy: 0.7349\n",
            "Epoch 106/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2177 - accuracy: 0.9234 - val_loss: 0.8041 - val_accuracy: 0.7952\n",
            "Epoch 107/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2337 - accuracy: 0.9225 - val_loss: 0.6129 - val_accuracy: 0.8193\n",
            "Epoch 108/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2025 - accuracy: 0.9263 - val_loss: 0.8539 - val_accuracy: 0.7831\n",
            "Epoch 109/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2104 - accuracy: 0.9292 - val_loss: 0.6432 - val_accuracy: 0.8193\n",
            "Epoch 110/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2266 - accuracy: 0.9206 - val_loss: 0.7555 - val_accuracy: 0.7952\n",
            "Epoch 111/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2713 - accuracy: 0.8986 - val_loss: 0.6717 - val_accuracy: 0.8313\n",
            "Epoch 112/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2202 - accuracy: 0.9206 - val_loss: 0.7632 - val_accuracy: 0.8072\n",
            "Epoch 113/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2255 - accuracy: 0.9244 - val_loss: 0.8744 - val_accuracy: 0.8072\n",
            "Epoch 114/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2207 - accuracy: 0.9215 - val_loss: 0.5591 - val_accuracy: 0.8313\n",
            "Epoch 115/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2016 - accuracy: 0.9254 - val_loss: 0.7086 - val_accuracy: 0.8313\n",
            "Epoch 116/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2259 - accuracy: 0.9254 - val_loss: 0.6346 - val_accuracy: 0.7952\n",
            "Epoch 117/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2203 - accuracy: 0.9263 - val_loss: 0.7694 - val_accuracy: 0.7711\n",
            "Epoch 118/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1855 - accuracy: 0.9349 - val_loss: 0.6861 - val_accuracy: 0.8072\n",
            "Epoch 119/300\n",
            "33/33 [==============================] - 15s 469ms/step - loss: 0.2177 - accuracy: 0.9167 - val_loss: 0.8787 - val_accuracy: 0.8313\n",
            "Epoch 120/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2018 - accuracy: 0.9368 - val_loss: 0.6488 - val_accuracy: 0.8193\n",
            "Epoch 121/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2282 - accuracy: 0.9263 - val_loss: 0.8419 - val_accuracy: 0.8072\n",
            "Epoch 122/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2159 - accuracy: 0.9167 - val_loss: 0.9396 - val_accuracy: 0.8313\n",
            "Epoch 123/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.1855 - accuracy: 0.9225 - val_loss: 0.5017 - val_accuracy: 0.8434\n",
            "Epoch 124/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2025 - accuracy: 0.9244 - val_loss: 0.7191 - val_accuracy: 0.8193\n",
            "Epoch 125/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2086 - accuracy: 0.9225 - val_loss: 0.7219 - val_accuracy: 0.7952\n",
            "Epoch 126/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1905 - accuracy: 0.9407 - val_loss: 0.9283 - val_accuracy: 0.8072\n",
            "Epoch 127/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2342 - accuracy: 0.9129 - val_loss: 0.7310 - val_accuracy: 0.7831\n",
            "Epoch 128/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2107 - accuracy: 0.9215 - val_loss: 0.6471 - val_accuracy: 0.8313\n",
            "Epoch 129/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2199 - accuracy: 0.9234 - val_loss: 0.7280 - val_accuracy: 0.7711\n",
            "Epoch 130/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2206 - accuracy: 0.9215 - val_loss: 0.8468 - val_accuracy: 0.8313\n",
            "Epoch 131/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.2070 - accuracy: 0.9263 - val_loss: 0.7964 - val_accuracy: 0.7952\n",
            "Epoch 132/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1978 - accuracy: 0.9244 - val_loss: 0.6844 - val_accuracy: 0.8072\n",
            "Epoch 133/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1716 - accuracy: 0.9407 - val_loss: 0.8780 - val_accuracy: 0.8072\n",
            "Epoch 134/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.1885 - accuracy: 0.9234 - val_loss: 0.7556 - val_accuracy: 0.8072\n",
            "Epoch 135/300\n",
            "33/33 [==============================] - 15s 456ms/step - loss: 0.1826 - accuracy: 0.9388 - val_loss: 0.7998 - val_accuracy: 0.8434\n",
            "Epoch 136/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.1742 - accuracy: 0.9349 - val_loss: 1.1382 - val_accuracy: 0.7952\n",
            "Epoch 137/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2205 - accuracy: 0.9148 - val_loss: 0.7875 - val_accuracy: 0.8193\n",
            "Epoch 138/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2056 - accuracy: 0.9321 - val_loss: 0.7456 - val_accuracy: 0.8193\n",
            "Epoch 139/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1922 - accuracy: 0.9282 - val_loss: 0.5906 - val_accuracy: 0.8072\n",
            "Epoch 140/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2067 - accuracy: 0.9187 - val_loss: 0.9038 - val_accuracy: 0.7952\n",
            "Epoch 141/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.1865 - accuracy: 0.9397 - val_loss: 0.8146 - val_accuracy: 0.7952\n",
            "Epoch 142/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2251 - accuracy: 0.9234 - val_loss: 0.8504 - val_accuracy: 0.7711\n",
            "Epoch 143/300\n",
            "33/33 [==============================] - 15s 457ms/step - loss: 0.2150 - accuracy: 0.9234 - val_loss: 0.7998 - val_accuracy: 0.7831\n",
            "Epoch 144/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2066 - accuracy: 0.9321 - val_loss: 0.7980 - val_accuracy: 0.8072\n",
            "Epoch 145/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1794 - accuracy: 0.9359 - val_loss: 0.9706 - val_accuracy: 0.8072\n",
            "Epoch 146/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1888 - accuracy: 0.9301 - val_loss: 0.5947 - val_accuracy: 0.8193\n",
            "Epoch 147/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2246 - accuracy: 0.9234 - val_loss: 0.5856 - val_accuracy: 0.8434\n",
            "Epoch 148/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2114 - accuracy: 0.9225 - val_loss: 0.5548 - val_accuracy: 0.8313\n",
            "Epoch 149/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1875 - accuracy: 0.9311 - val_loss: 0.6122 - val_accuracy: 0.7952\n",
            "Epoch 150/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2089 - accuracy: 0.9301 - val_loss: 0.6512 - val_accuracy: 0.7952\n",
            "Epoch 151/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1762 - accuracy: 0.9292 - val_loss: 0.5541 - val_accuracy: 0.8313\n",
            "Epoch 152/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2026 - accuracy: 0.9263 - val_loss: 0.7295 - val_accuracy: 0.8193\n",
            "Epoch 153/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2160 - accuracy: 0.9158 - val_loss: 0.7926 - val_accuracy: 0.7831\n",
            "Epoch 154/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.2103 - accuracy: 0.9196 - val_loss: 0.6798 - val_accuracy: 0.8313\n",
            "Epoch 155/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2052 - accuracy: 0.9244 - val_loss: 0.6664 - val_accuracy: 0.8072\n",
            "Epoch 156/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2071 - accuracy: 0.9206 - val_loss: 0.7378 - val_accuracy: 0.7831\n",
            "Epoch 157/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2369 - accuracy: 0.9263 - val_loss: 0.9085 - val_accuracy: 0.7952\n",
            "Epoch 158/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2104 - accuracy: 0.9187 - val_loss: 1.0019 - val_accuracy: 0.7831\n",
            "Epoch 159/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1894 - accuracy: 0.9349 - val_loss: 0.5915 - val_accuracy: 0.8193\n",
            "Epoch 160/300\n",
            "33/33 [==============================] - 16s 470ms/step - loss: 0.1895 - accuracy: 0.9435 - val_loss: 0.6545 - val_accuracy: 0.8554\n",
            "Epoch 161/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1939 - accuracy: 0.9273 - val_loss: 0.6977 - val_accuracy: 0.8313\n",
            "Epoch 162/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.2057 - accuracy: 0.9215 - val_loss: 0.7717 - val_accuracy: 0.8072\n",
            "Epoch 163/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1895 - accuracy: 0.9340 - val_loss: 0.6928 - val_accuracy: 0.7952\n",
            "Epoch 164/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1996 - accuracy: 0.9330 - val_loss: 0.8134 - val_accuracy: 0.7952\n",
            "Epoch 165/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2167 - accuracy: 0.9225 - val_loss: 1.0218 - val_accuracy: 0.7831\n",
            "Epoch 166/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1988 - accuracy: 0.9273 - val_loss: 0.6622 - val_accuracy: 0.8434\n",
            "Epoch 167/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1751 - accuracy: 0.9330 - val_loss: 0.7551 - val_accuracy: 0.7831\n",
            "Epoch 168/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2077 - accuracy: 0.9273 - val_loss: 0.5918 - val_accuracy: 0.8313\n",
            "Epoch 169/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2037 - accuracy: 0.9359 - val_loss: 0.8330 - val_accuracy: 0.8072\n",
            "Epoch 170/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1842 - accuracy: 0.9349 - val_loss: 0.8946 - val_accuracy: 0.8072\n",
            "Epoch 171/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1929 - accuracy: 0.9321 - val_loss: 0.6786 - val_accuracy: 0.8193\n",
            "Epoch 172/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1896 - accuracy: 0.9359 - val_loss: 0.6341 - val_accuracy: 0.8434\n",
            "Epoch 173/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1934 - accuracy: 0.9321 - val_loss: 0.7881 - val_accuracy: 0.8434\n",
            "Epoch 174/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2254 - accuracy: 0.9263 - val_loss: 0.5365 - val_accuracy: 0.8313\n",
            "Epoch 175/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2065 - accuracy: 0.9225 - val_loss: 0.6709 - val_accuracy: 0.8072\n",
            "Epoch 176/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2060 - accuracy: 0.9321 - val_loss: 0.8686 - val_accuracy: 0.8072\n",
            "Epoch 177/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1818 - accuracy: 0.9368 - val_loss: 0.7195 - val_accuracy: 0.8313\n",
            "Epoch 178/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1847 - accuracy: 0.9330 - val_loss: 0.6206 - val_accuracy: 0.8554\n",
            "Epoch 179/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1871 - accuracy: 0.9292 - val_loss: 0.6258 - val_accuracy: 0.8193\n",
            "Epoch 180/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.1903 - accuracy: 0.9292 - val_loss: 0.9581 - val_accuracy: 0.7952\n",
            "Epoch 181/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1968 - accuracy: 0.9321 - val_loss: 0.7678 - val_accuracy: 0.8313\n",
            "Epoch 182/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2051 - accuracy: 0.9234 - val_loss: 0.5581 - val_accuracy: 0.8313\n",
            "Epoch 183/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1933 - accuracy: 0.9321 - val_loss: 0.7308 - val_accuracy: 0.8313\n",
            "Epoch 184/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1964 - accuracy: 0.9254 - val_loss: 0.8170 - val_accuracy: 0.8193\n",
            "Epoch 185/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2306 - accuracy: 0.9263 - val_loss: 1.0380 - val_accuracy: 0.7349\n",
            "Epoch 186/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1954 - accuracy: 0.9282 - val_loss: 0.7468 - val_accuracy: 0.7831\n",
            "Epoch 187/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2053 - accuracy: 0.9292 - val_loss: 0.9905 - val_accuracy: 0.7831\n",
            "Epoch 188/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.1738 - accuracy: 0.9435 - val_loss: 0.7819 - val_accuracy: 0.8313\n",
            "Epoch 189/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1816 - accuracy: 0.9368 - val_loss: 0.7380 - val_accuracy: 0.7952\n",
            "Epoch 190/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1798 - accuracy: 0.9378 - val_loss: 0.7378 - val_accuracy: 0.7952\n",
            "Epoch 191/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1947 - accuracy: 0.9254 - val_loss: 0.6745 - val_accuracy: 0.8193\n",
            "Epoch 192/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1788 - accuracy: 0.9368 - val_loss: 0.9600 - val_accuracy: 0.7952\n",
            "Epoch 193/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2012 - accuracy: 0.9244 - val_loss: 0.8598 - val_accuracy: 0.8072\n",
            "Epoch 194/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2092 - accuracy: 0.9234 - val_loss: 0.7501 - val_accuracy: 0.7711\n",
            "Epoch 195/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1624 - accuracy: 0.9388 - val_loss: 0.7632 - val_accuracy: 0.8072\n",
            "Epoch 196/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.2226 - accuracy: 0.9167 - val_loss: 0.7306 - val_accuracy: 0.8072\n",
            "Epoch 197/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.1953 - accuracy: 0.9349 - val_loss: 0.8518 - val_accuracy: 0.8313\n",
            "Epoch 198/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2071 - accuracy: 0.9292 - val_loss: 0.9991 - val_accuracy: 0.7711\n",
            "Epoch 199/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1793 - accuracy: 0.9254 - val_loss: 0.8189 - val_accuracy: 0.7711\n",
            "Epoch 200/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2081 - accuracy: 0.9340 - val_loss: 0.6674 - val_accuracy: 0.8313\n",
            "Epoch 201/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2095 - accuracy: 0.9244 - val_loss: 0.8033 - val_accuracy: 0.7952\n",
            "Epoch 202/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2145 - accuracy: 0.9330 - val_loss: 0.7574 - val_accuracy: 0.8434\n",
            "Epoch 203/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1769 - accuracy: 0.9407 - val_loss: 0.8609 - val_accuracy: 0.8313\n",
            "Epoch 204/300\n",
            "33/33 [==============================] - 15s 459ms/step - loss: 0.1767 - accuracy: 0.9359 - val_loss: 0.8108 - val_accuracy: 0.8072\n",
            "Epoch 205/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1934 - accuracy: 0.9273 - val_loss: 0.7898 - val_accuracy: 0.8193\n",
            "Epoch 206/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1988 - accuracy: 0.9311 - val_loss: 0.7739 - val_accuracy: 0.8193\n",
            "Epoch 207/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2083 - accuracy: 0.9292 - val_loss: 0.8497 - val_accuracy: 0.8193\n",
            "Epoch 208/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2125 - accuracy: 0.9340 - val_loss: 0.6380 - val_accuracy: 0.7711\n",
            "Epoch 209/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2076 - accuracy: 0.9187 - val_loss: 0.8609 - val_accuracy: 0.7952\n",
            "Epoch 210/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2008 - accuracy: 0.9330 - val_loss: 0.9049 - val_accuracy: 0.8193\n",
            "Epoch 211/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.1905 - accuracy: 0.9292 - val_loss: 0.7811 - val_accuracy: 0.7831\n",
            "Epoch 212/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1900 - accuracy: 0.9349 - val_loss: 0.7776 - val_accuracy: 0.8072\n",
            "Epoch 213/300\n",
            "33/33 [==============================] - 15s 469ms/step - loss: 0.1608 - accuracy: 0.9359 - val_loss: 0.8347 - val_accuracy: 0.7711\n",
            "Epoch 214/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1943 - accuracy: 0.9301 - val_loss: 0.7926 - val_accuracy: 0.8072\n",
            "Epoch 215/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.1741 - accuracy: 0.9359 - val_loss: 0.9144 - val_accuracy: 0.8072\n",
            "Epoch 216/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1749 - accuracy: 0.9435 - val_loss: 0.7929 - val_accuracy: 0.8193\n",
            "Epoch 217/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.2173 - accuracy: 0.9225 - val_loss: 0.8783 - val_accuracy: 0.7952\n",
            "Epoch 218/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1643 - accuracy: 0.9416 - val_loss: 0.7974 - val_accuracy: 0.8072\n",
            "Epoch 219/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1866 - accuracy: 0.9368 - val_loss: 0.6556 - val_accuracy: 0.8193\n",
            "Epoch 220/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1568 - accuracy: 0.9512 - val_loss: 0.7997 - val_accuracy: 0.8072\n",
            "Epoch 221/300\n",
            "33/33 [==============================] - 16s 469ms/step - loss: 0.1936 - accuracy: 0.9301 - val_loss: 1.0949 - val_accuracy: 0.7831\n",
            "Epoch 222/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1491 - accuracy: 0.9522 - val_loss: 0.8432 - val_accuracy: 0.8313\n",
            "Epoch 223/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.1909 - accuracy: 0.9292 - val_loss: 0.5644 - val_accuracy: 0.8313\n",
            "Epoch 224/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1635 - accuracy: 0.9426 - val_loss: 0.7432 - val_accuracy: 0.8193\n",
            "Epoch 225/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1898 - accuracy: 0.9301 - val_loss: 0.8339 - val_accuracy: 0.7831\n",
            "Epoch 226/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1784 - accuracy: 0.9349 - val_loss: 0.5466 - val_accuracy: 0.8554\n",
            "Epoch 227/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1663 - accuracy: 0.9368 - val_loss: 0.7124 - val_accuracy: 0.8795\n",
            "Epoch 228/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1758 - accuracy: 0.9340 - val_loss: 0.9142 - val_accuracy: 0.7952\n",
            "Epoch 229/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.2015 - accuracy: 0.9215 - val_loss: 0.5553 - val_accuracy: 0.8193\n",
            "Epoch 230/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2067 - accuracy: 0.9301 - val_loss: 0.8201 - val_accuracy: 0.7952\n",
            "Epoch 231/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1616 - accuracy: 0.9426 - val_loss: 0.8256 - val_accuracy: 0.8313\n",
            "Epoch 232/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1740 - accuracy: 0.9397 - val_loss: 0.5888 - val_accuracy: 0.8072\n",
            "Epoch 233/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1982 - accuracy: 0.9263 - val_loss: 1.1578 - val_accuracy: 0.7952\n",
            "Epoch 234/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1561 - accuracy: 0.9455 - val_loss: 0.7998 - val_accuracy: 0.7952\n",
            "Epoch 235/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1683 - accuracy: 0.9435 - val_loss: 0.6892 - val_accuracy: 0.8434\n",
            "Epoch 236/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1654 - accuracy: 0.9368 - val_loss: 0.8340 - val_accuracy: 0.7711\n",
            "Epoch 237/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1538 - accuracy: 0.9397 - val_loss: 0.6228 - val_accuracy: 0.8193\n",
            "Epoch 238/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1866 - accuracy: 0.9330 - val_loss: 0.7317 - val_accuracy: 0.8193\n",
            "Epoch 239/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1870 - accuracy: 0.9359 - val_loss: 0.8118 - val_accuracy: 0.8072\n",
            "Epoch 240/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1974 - accuracy: 0.9292 - val_loss: 0.5939 - val_accuracy: 0.7952\n",
            "Epoch 241/300\n",
            "33/33 [==============================] - 16s 472ms/step - loss: 0.1732 - accuracy: 0.9349 - val_loss: 0.5228 - val_accuracy: 0.8554\n",
            "Epoch 242/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1654 - accuracy: 0.9388 - val_loss: 0.8802 - val_accuracy: 0.7831\n",
            "Epoch 243/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1613 - accuracy: 0.9388 - val_loss: 0.5841 - val_accuracy: 0.8554\n",
            "Epoch 244/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1868 - accuracy: 0.9416 - val_loss: 0.7447 - val_accuracy: 0.8434\n",
            "Epoch 245/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1688 - accuracy: 0.9416 - val_loss: 0.8055 - val_accuracy: 0.7831\n",
            "Epoch 246/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1948 - accuracy: 0.9254 - val_loss: 0.6645 - val_accuracy: 0.7952\n",
            "Epoch 247/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1964 - accuracy: 0.9330 - val_loss: 0.7952 - val_accuracy: 0.8193\n",
            "Epoch 248/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1767 - accuracy: 0.9368 - val_loss: 0.6417 - val_accuracy: 0.8313\n",
            "Epoch 249/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.2071 - accuracy: 0.9254 - val_loss: 0.9300 - val_accuracy: 0.7831\n",
            "Epoch 250/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.1830 - accuracy: 0.9407 - val_loss: 0.7610 - val_accuracy: 0.8313\n",
            "Epoch 251/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1344 - accuracy: 0.9455 - val_loss: 0.6099 - val_accuracy: 0.8675\n",
            "Epoch 252/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1578 - accuracy: 0.9388 - val_loss: 0.8469 - val_accuracy: 0.8072\n",
            "Epoch 253/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.2074 - accuracy: 0.9368 - val_loss: 0.8665 - val_accuracy: 0.7952\n",
            "Epoch 254/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1936 - accuracy: 0.9234 - val_loss: 0.7392 - val_accuracy: 0.8554\n",
            "Epoch 255/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1794 - accuracy: 0.9378 - val_loss: 1.0918 - val_accuracy: 0.8072\n",
            "Epoch 256/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.2105 - accuracy: 0.9282 - val_loss: 0.4974 - val_accuracy: 0.8554\n",
            "Epoch 257/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1783 - accuracy: 0.9349 - val_loss: 1.0241 - val_accuracy: 0.7590\n",
            "Epoch 258/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.2338 - accuracy: 0.9215 - val_loss: 0.9730 - val_accuracy: 0.7590\n",
            "Epoch 259/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1859 - accuracy: 0.9397 - val_loss: 0.5313 - val_accuracy: 0.8313\n",
            "Epoch 260/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1611 - accuracy: 0.9378 - val_loss: 0.8781 - val_accuracy: 0.7952\n",
            "Epoch 261/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.1684 - accuracy: 0.9397 - val_loss: 0.7032 - val_accuracy: 0.8193\n",
            "Epoch 262/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1570 - accuracy: 0.9416 - val_loss: 0.9054 - val_accuracy: 0.8072\n",
            "Epoch 263/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.1881 - accuracy: 0.9244 - val_loss: 0.7137 - val_accuracy: 0.8193\n",
            "Epoch 264/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1959 - accuracy: 0.9234 - val_loss: 1.0463 - val_accuracy: 0.7470\n",
            "Epoch 265/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1851 - accuracy: 0.9321 - val_loss: 0.7828 - val_accuracy: 0.8313\n",
            "Epoch 266/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1783 - accuracy: 0.9273 - val_loss: 0.9058 - val_accuracy: 0.8434\n",
            "Epoch 267/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1752 - accuracy: 0.9368 - val_loss: 0.5871 - val_accuracy: 0.8434\n",
            "Epoch 268/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1735 - accuracy: 0.9311 - val_loss: 0.7287 - val_accuracy: 0.8313\n",
            "Epoch 269/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1917 - accuracy: 0.9234 - val_loss: 1.0807 - val_accuracy: 0.7831\n",
            "Epoch 270/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1910 - accuracy: 0.9311 - val_loss: 0.6949 - val_accuracy: 0.8193\n",
            "Epoch 271/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1829 - accuracy: 0.9340 - val_loss: 0.7828 - val_accuracy: 0.7831\n",
            "Epoch 272/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1704 - accuracy: 0.9388 - val_loss: 0.8939 - val_accuracy: 0.7831\n",
            "Epoch 273/300\n",
            "33/33 [==============================] - 15s 461ms/step - loss: 0.1803 - accuracy: 0.9416 - val_loss: 0.8823 - val_accuracy: 0.8193\n",
            "Epoch 274/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.1864 - accuracy: 0.9397 - val_loss: 0.6626 - val_accuracy: 0.8072\n",
            "Epoch 275/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1638 - accuracy: 0.9435 - val_loss: 0.7378 - val_accuracy: 0.8193\n",
            "Epoch 276/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1813 - accuracy: 0.9397 - val_loss: 0.8319 - val_accuracy: 0.7711\n",
            "Epoch 277/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1653 - accuracy: 0.9445 - val_loss: 0.8201 - val_accuracy: 0.7952\n",
            "Epoch 278/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.1577 - accuracy: 0.9483 - val_loss: 0.7473 - val_accuracy: 0.8072\n",
            "Epoch 279/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1736 - accuracy: 0.9388 - val_loss: 0.9182 - val_accuracy: 0.7590\n",
            "Epoch 280/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1799 - accuracy: 0.9435 - val_loss: 0.7150 - val_accuracy: 0.8434\n",
            "Epoch 281/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1769 - accuracy: 0.9397 - val_loss: 0.7163 - val_accuracy: 0.8072\n",
            "Epoch 282/300\n",
            "33/33 [==============================] - 15s 468ms/step - loss: 0.2194 - accuracy: 0.9120 - val_loss: 0.8250 - val_accuracy: 0.7711\n",
            "Epoch 283/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1698 - accuracy: 0.9388 - val_loss: 0.7884 - val_accuracy: 0.8072\n",
            "Epoch 284/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1620 - accuracy: 0.9359 - val_loss: 0.7158 - val_accuracy: 0.7590\n",
            "Epoch 285/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1816 - accuracy: 0.9311 - val_loss: 0.8851 - val_accuracy: 0.8554\n",
            "Epoch 286/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.1881 - accuracy: 0.9349 - val_loss: 0.7938 - val_accuracy: 0.8554\n",
            "Epoch 287/300\n",
            "33/33 [==============================] - 15s 458ms/step - loss: 0.2060 - accuracy: 0.9263 - val_loss: 0.9498 - val_accuracy: 0.8072\n",
            "Epoch 288/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1774 - accuracy: 0.9416 - val_loss: 0.8267 - val_accuracy: 0.7952\n",
            "Epoch 289/300\n",
            "33/33 [==============================] - 15s 463ms/step - loss: 0.2088 - accuracy: 0.9234 - val_loss: 0.6756 - val_accuracy: 0.8193\n",
            "Epoch 290/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1714 - accuracy: 0.9311 - val_loss: 0.6884 - val_accuracy: 0.8313\n",
            "Epoch 291/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1824 - accuracy: 0.9225 - val_loss: 0.8437 - val_accuracy: 0.7831\n",
            "Epoch 292/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1778 - accuracy: 0.9349 - val_loss: 0.6115 - val_accuracy: 0.8554\n",
            "Epoch 293/300\n",
            "33/33 [==============================] - 15s 464ms/step - loss: 0.1595 - accuracy: 0.9388 - val_loss: 0.9224 - val_accuracy: 0.8072\n",
            "Epoch 294/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1826 - accuracy: 0.9311 - val_loss: 0.5838 - val_accuracy: 0.8313\n",
            "Epoch 295/300\n",
            "33/33 [==============================] - 15s 466ms/step - loss: 0.1823 - accuracy: 0.9340 - val_loss: 0.7572 - val_accuracy: 0.8554\n",
            "Epoch 296/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1535 - accuracy: 0.9455 - val_loss: 0.7075 - val_accuracy: 0.8313\n",
            "Epoch 297/300\n",
            "33/33 [==============================] - 15s 460ms/step - loss: 0.1786 - accuracy: 0.9388 - val_loss: 0.8585 - val_accuracy: 0.8313\n",
            "Epoch 298/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1739 - accuracy: 0.9359 - val_loss: 0.4840 - val_accuracy: 0.8795\n",
            "Epoch 299/300\n",
            "33/33 [==============================] - 15s 465ms/step - loss: 0.1667 - accuracy: 0.9407 - val_loss: 0.8796 - val_accuracy: 0.7952\n",
            "Epoch 300/300\n",
            "33/33 [==============================] - 15s 462ms/step - loss: 0.1752 - accuracy: 0.9407 - val_loss: 0.7809 - val_accuracy: 0.8193\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "A0mXAx0A3iny",
        "outputId": "a9670767-c88b-4553-d557-956c28177edf"
      },
      "source": [
        "final_loss, final_accuracy = model.evaluate(valid_generator)\r\n",
        "print(\"Final loss: {:.2f}\".format(final_loss))\r\n",
        "print(\"Final Valid accuracy: {:.2f}%\".format(final_accuracy * 100))\r\n",
        "\r\n",
        "final_loss, final_accuracy = model.evaluate(train_generator)\r\n",
        "print(\"Final loss: {:.2f}\".format(final_loss))\r\n",
        "print(\"Final Train accuracy: {:.2f}%\".format(final_accuracy * 100))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "3/3 [==============================] - 1s 473ms/step - loss: 0.7212 - accuracy: 0.8193\n",
            "Final loss: 0.72\n",
            "Final Valid accuracy: 81.93%\n",
            "33/33 [==============================] - 14s 414ms/step - loss: 0.0791 - accuracy: 0.9761\n",
            "Final loss: 0.08\n",
            "Final Train accuracy: 97.61%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J6QU-xpcAM8k"
      },
      "source": [
        "model.save('/content/gdrive/MyDrive/Eye Diseases/eye_disease_detection.h5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8LDasYqrEiNk",
        "outputId": "0ecdfe89-d631-4239-f4da-82afce27153a"
      },
      "source": [
        "import numpy as np\r\n",
        "\r\n",
        "val_image_batch, val_label_batch = next(iter(valid_generator))\r\n",
        "true_label_ids = np.argmax(val_label_batch, axis=-1)\r\n",
        "print(\"Validation batch shape:\", val_image_batch.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Validation batch shape: (32, 200, 200, 3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "maA0EaynFLVA",
        "outputId": "2def9a31-89d5-449c-df41-5be743613d1a"
      },
      "source": [
        "dataset_labels = sorted(train_generator.class_indices.items(), key=lambda pair:pair[1])\r\n",
        "dataset_labels = np.array([key.title() for key, value in dataset_labels])\r\n",
        "print(dataset_labels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['Arcus Senilis' 'Cataracts' 'Endophthalmitis' 'Hemangioma' 'Hyphema']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SB02IUBS4ZGH"
      },
      "source": [
        "Predicting on validation set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Y2AqhN_sFUfk",
        "outputId": "308fc3ee-f5e7-47cb-b46b-abb1cf3ddd1f"
      },
      "source": [
        "tf_model_predictions = model.predict(val_image_batch)\r\n",
        "print(\"Prediction results shape:\", tf_model_predictions.shape)\r\n",
        "print(tf_model_predictions)\r\n",
        "predicted_ids = np.argmax(tf_model_predictions, axis=-1)\r\n",
        "predicted_labels = dataset_labels[predicted_ids]\r\n",
        "print(predicted_labels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Prediction results shape: (32, 5)\n",
            "[[5.30476719e-02 9.45869148e-01 7.01819663e-05 4.88160003e-05\n",
            "  9.64164559e-04]\n",
            " [1.00943704e-04 9.99897122e-01 8.11278227e-08 1.24229132e-07\n",
            "  1.70897761e-06]\n",
            " [1.02753507e-03 6.63300161e-05 9.81068671e-01 2.44020222e-04\n",
            "  1.75935514e-02]\n",
            " [5.15712500e-02 7.63550282e-01 8.00942816e-03 4.22490615e-04\n",
            "  1.76446497e-01]\n",
            " [1.75755355e-04 3.66787463e-05 7.14716196e-01 1.18185075e-04\n",
            "  2.84953296e-01]\n",
            " [5.64435720e-01 1.77410850e-03 2.13300407e-01 6.23232918e-03\n",
            "  2.14257389e-01]\n",
            " [1.40494012e-04 4.08467276e-05 1.41212333e-03 8.69257492e-04\n",
            "  9.97537255e-01]\n",
            " [1.11508009e-03 5.07826335e-05 8.39429922e-05 9.98690069e-01\n",
            "  6.01527354e-05]\n",
            " [1.15322627e-01 3.46851423e-02 6.20426536e-02 9.29986127e-03\n",
            "  7.78649688e-01]\n",
            " [5.74010483e-04 4.90027560e-05 4.97739129e-05 5.78649851e-05\n",
            "  9.99269307e-01]\n",
            " [9.83977854e-01 1.51552129e-04 1.57604776e-02 1.13339902e-05\n",
            "  9.87565145e-05]\n",
            " [1.01009607e-02 9.84148264e-01 4.15319623e-03 8.09078992e-05\n",
            "  1.51675765e-03]\n",
            " [2.75353455e-06 7.99355007e-07 7.34769419e-05 6.53413338e-07\n",
            "  9.99922276e-01]\n",
            " [1.53307956e-05 9.99984384e-01 5.65977398e-09 2.47120013e-09\n",
            "  2.02876237e-07]\n",
            " [9.99956012e-01 2.18080459e-05 9.94039851e-07 3.53994977e-07\n",
            "  2.08253623e-05]\n",
            " [6.52790785e-01 3.01170081e-01 1.95259380e-03 9.19269572e-04\n",
            "  4.31672297e-02]\n",
            " [6.71492191e-04 5.04473828e-05 9.97629344e-01 4.17348529e-05\n",
            "  1.60703156e-03]\n",
            " [7.15645874e-05 3.54704275e-06 2.01816101e-06 1.75833816e-06\n",
            "  9.99921083e-01]\n",
            " [4.36362438e-03 1.61945471e-04 9.86326814e-01 3.12894263e-04\n",
            "  8.83468334e-03]\n",
            " [9.99998212e-01 1.09576000e-07 1.31164870e-06 8.48713597e-08\n",
            "  1.78962622e-07]\n",
            " [9.99982119e-01 8.61032277e-06 3.85197723e-07 4.17624449e-07\n",
            "  8.44017632e-06]\n",
            " [9.97987509e-01 1.35531661e-03 1.28977961e-04 4.17775591e-04\n",
            "  1.10368674e-04]\n",
            " [3.29736084e-01 6.68428004e-01 4.14943752e-05 3.97625954e-05\n",
            "  1.75458961e-03]\n",
            " [6.52034432e-02 2.64887756e-04 2.92954122e-04 9.68767126e-05\n",
            "  9.34141874e-01]\n",
            " [6.78103138e-03 9.90779085e-07 9.93116856e-01 5.18973729e-06\n",
            "  9.59446625e-05]\n",
            " [9.94026065e-01 8.52788624e-04 2.92769237e-03 3.92843394e-05\n",
            "  2.15421617e-03]\n",
            " [4.31315915e-04 9.99564111e-01 2.06448803e-07 5.83512438e-08\n",
            "  4.31597300e-06]\n",
            " [9.37987636e-07 9.99999046e-01 1.28229455e-10 1.76002546e-10\n",
            "  1.68443670e-09]\n",
            " [2.05691744e-04 3.59092815e-06 6.63162573e-05 2.82919768e-06\n",
            "  9.99721587e-01]\n",
            " [9.98680413e-01 6.91961213e-06 1.24969403e-03 3.19406900e-05\n",
            "  3.10854884e-05]\n",
            " [3.62244528e-03 1.71457836e-03 9.67904449e-01 1.62998127e-04\n",
            "  2.65954621e-02]\n",
            " [9.99387503e-01 5.46817500e-05 3.70990048e-04 9.63235798e-05\n",
            "  9.04279223e-05]]\n",
            "['Cataracts' 'Cataracts' 'Endophthalmitis' 'Cataracts' 'Endophthalmitis'\n",
            " 'Arcus Senilis' 'Hyphema' 'Hemangioma' 'Hyphema' 'Hyphema'\n",
            " 'Arcus Senilis' 'Cataracts' 'Hyphema' 'Cataracts' 'Arcus Senilis'\n",
            " 'Arcus Senilis' 'Endophthalmitis' 'Hyphema' 'Endophthalmitis'\n",
            " 'Arcus Senilis' 'Arcus Senilis' 'Arcus Senilis' 'Cataracts' 'Hyphema'\n",
            " 'Endophthalmitis' 'Arcus Senilis' 'Cataracts' 'Cataracts' 'Hyphema'\n",
            " 'Arcus Senilis' 'Endophthalmitis' 'Arcus Senilis']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WOtZjn2y4cOq"
      },
      "source": [
        "Validation Set Prediction Results: "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 585
        },
        "id": "-LiouQo9Fe3i",
        "outputId": "1e121843-e12a-425f-ef27-bd4fc85ccf4e"
      },
      "source": [
        "import matplotlib.pyplot as plt\r\n",
        "plt.figure(figsize=(10,9))\r\n",
        "plt.subplots_adjust(hspace=0.5)\r\n",
        "for n in range((len(predicted_labels)-2)):\r\n",
        "  plt.subplot(6,5,n+1)\r\n",
        "  plt.imshow(val_image_batch[n])\r\n",
        "  color = \"green\" if predicted_ids[n] == true_label_ids[n] else \"red\"\r\n",
        "  plt.title(predicted_labels[n].title(), color=color)\r\n",
        "  plt.axis('off')\r\n",
        "_ = plt.suptitle(\"Model predictions (green: correct, red: incorrect)\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x648 with 30 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tGE0YWdBIlWB"
      },
      "source": [
        "Model Testing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tcOZ-IfbINUT"
      },
      "source": [
        "from keras.preprocessing.image import load_img\r\n",
        " \r\n",
        "image = load_img('/content/gdrive/MyDrive/Eye Diseases/Test Images/_185_capillaryhemangioma1.jpg', target_size=(200,200))\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6iW7jMkpLfkv"
      },
      "source": [
        "from keras.preprocessing.image import img_to_array\r\n",
        "# convert the image pixels to a numpy array\r\n",
        "image = img_to_array(image)\r\n",
        "#image = np.repeat(image, 3, 2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bQeQ5MCPLk_g",
        "outputId": "b193d778-6bb6-4ee5-9c86-06c1b2cbf654"
      },
      "source": [
        "image = np.expand_dims(image,axis=0)\r\n",
        "image = np.array(image)/255.0\r\n",
        "print(image.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(1, 200, 200, 3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GDPMwl1gLnY-",
        "outputId": "04e55426-0529-4dc3-ef0a-c5f2bb3091db"
      },
      "source": [
        "pred = model.predict(image)\r\n",
        "print(pred.shape)\r\n",
        "print (pred)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(1, 5)\n",
            "[[0.2072205  0.00075663 0.03469858 0.71819454 0.03912976]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GyyDWkwJPAxH",
        "outputId": "54693911-d2ea-47b8-bb28-fea92f2ab8b5"
      },
      "source": [
        "predicted_id = np.argmax(pred, axis=-1)\r\n",
        "predicted_labels = dataset_labels[predicted_id]\r\n",
        "print(predicted_labels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['Hemangioma']\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}