vit_l32.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "machine_shape": "hm",
      "gpuType": "T4",
      "cell_execution_strategy": "setup",
      "authorship_tag": "ABX9TyNA3v60UZpwoQg7A2zOLHK1",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Crepopcorn/cat-vs-dog-vit-project/blob/main/vit_l32_scratch.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### **Dogs-vs-Cats classifier (vit_132)**\n"
      ],
      "metadata": {
        "id": "BJOcpoBYkSnI"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**1) Pre-load module**"
      ],
      "metadata": {
        "id": "0HCrJhB4kc6G"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JMYBQt_qkLvr",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a5ebe411-301a-45ae-a6b6-27c98d837861"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting tensorflow-addons\n",
            "  Downloading tensorflow_addons-0.22.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (612 kB)\n",
            "\u001b[?25l     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/612.3 kB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r\u001b[2K     \u001b[91m━━━━━━━━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m235.5/612.3 kB\u001b[0m \u001b[31m6.9 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m612.3/612.3 kB\u001b[0m \u001b[31m11.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: packaging in /usr/local/lib/python3.10/dist-packages (from tensorflow-addons) (23.2)\n",
            "Collecting typeguard<3.0.0,>=2.7 (from tensorflow-addons)\n",
            "  Downloading typeguard-2.13.3-py3-none-any.whl (17 kB)\n",
            "Installing collected packages: typeguard, tensorflow-addons\n",
            "Successfully installed tensorflow-addons-0.22.0 typeguard-2.13.3\n",
            "Collecting vit-keras\n",
            "  Downloading vit_keras-0.1.2-py3-none-any.whl (24 kB)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from vit-keras) (1.11.3)\n",
            "Collecting validators (from vit-keras)\n",
            "  Downloading validators-0.22.0-py3-none-any.whl (26 kB)\n",
            "Requirement already satisfied: numpy<1.28.0,>=1.21.6 in /usr/local/lib/python3.10/dist-packages (from scipy->vit-keras) (1.23.5)\n",
            "Installing collected packages: validators, vit-keras\n",
            "Successfully installed validators-0.22.0 vit-keras-0.1.2\n",
            "Tue Oct 31 14:45:45 2023       \n",
            "+-----------------------------------------------------------------------------+\n",
            "| NVIDIA-SMI 525.105.17   Driver Version: 525.105.17   CUDA Version: 12.0     |\n",
            "|-------------------------------+----------------------+----------------------+\n",
            "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
            "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
            "|                               |                      |               MIG M. |\n",
            "|===============================+======================+======================|\n",
            "|   0  Tesla T4            Off  | 00000000:00:04.0 Off |                    0 |\n",
            "| N/A   42C    P8    12W /  70W |      0MiB / 15360MiB |      0%      Default |\n",
            "|                               |                      |                  N/A |\n",
            "+-------------------------------+----------------------+----------------------+\n",
            "                                                                               \n",
            "+-----------------------------------------------------------------------------+\n",
            "| Processes:                                                                  |\n",
            "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
            "|        ID   ID                                                   Usage      |\n",
            "|=============================================================================|\n",
            "|  No running processes found                                                 |\n",
            "+-----------------------------------------------------------------------------+\n"
          ]
        }
      ],
      "source": [
        "!pip install tensorflow-addons\n",
        "!pip install vit-keras\n",
        "!nvidia-smi"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount(\"/content/drive/\")"
      ],
      "metadata": {
        "id": "KHWvSkp4sbpR",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a69fb3e2-c374-46a9-a3f2-192b0d582b91"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive/\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "img_size=[256,256]\n",
        "train_dir='/content/drive/My Drive/data/train'\n",
        "test_dir='/content/drive/My Drive/data/val'\n",
        "!ls '/content/drive/My Drive/data/train'"
      ],
      "metadata": {
        "id": "tIYkhmh1tRTc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "989ae9ec-9681-4c9c-8638-a15233dbb589"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "cat  dog\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from vit_keras import vit, utils\n",
        "from keras.models import Model\n",
        "from keras.layers import Dense, Dropout, Flatten\n",
        "import matplotlib.pyplot as plt\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "from keras.models import Sequential,Model\n",
        "#from keras.optimizers import Adam\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.optimizers.legacy import Adam"
      ],
      "metadata": {
        "id": "RAl_MMaztTpG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2474216a-92fa-414f-a554-ee550054df88"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/tensorflow_addons/utils/tfa_eol_msg.py:23: UserWarning: \n",
            "\n",
            "TensorFlow Addons (TFA) has ended development and introduction of new features.\n",
            "TFA has entered a minimal maintenance and release mode until a planned end of life in May 2024.\n",
            "Please modify downstream libraries to take dependencies from other repositories in our TensorFlow community (e.g. Keras, Keras-CV, and Keras-NLP). \n",
            "\n",
            "For more information see: https://github.com/tensorflow/addons/issues/2807 \n",
            "\n",
            "  warnings.warn(\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**2) Preprocessing**"
      ],
      "metadata": {
        "id": "X_tWBXbdvPQU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "image_size = 256\n",
        "base_model = vit.vit_l32(\n",
        "    image_size=image_size,\n",
        "    activation='relu',\n",
        "    pretrained=True,\n",
        "    include_top=False,\n",
        "    pretrained_top=False,\n",
        ")"
      ],
      "metadata": {
        "id": "zMTXKjvmtVcc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a2e77fc7-fb51-440d-fc4b-e9ee1e2cadbf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://github.com/faustomorales/vit-keras/releases/download/dl/ViT-L_32_imagenet21k+imagenet2012.npz\n",
            "1226658854/1226658854 [==============================] - 4s 0us/step\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/vit_keras/utils.py:81: UserWarning: Resizing position embeddings from 12, 12 to 8, 8\n",
            "  warnings.warn(\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(base_model.summary())"
      ],
      "metadata": {
        "id": "m4iMrTiVtX81",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9c630abd-8fa6-40de-c364-6e813d448e54"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"vit-l32\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 256, 256, 3)]     0         \n",
            "                                                                 \n",
            " embedding (Conv2D)          (None, 8, 8, 1024)        3146752   \n",
            "                                                                 \n",
            " reshape (Reshape)           (None, 64, 1024)          0         \n",
            "                                                                 \n",
            " class_token (ClassToken)    (None, 65, 1024)          1024      \n",
            "                                                                 \n",
            " Transformer/posembed_input  (None, 65, 1024)          66560     \n",
            "  (AddPositionEmbs)                                              \n",
            "                                                                 \n",
            " Transformer/encoderblock_0  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_3  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_4  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_5  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_6  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_7  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_8  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_9  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 0 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 1 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 2 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 3 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 4 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 5 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 6 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 7 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 8 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 9 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 0 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 1 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 2 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 3 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoder_norm (  (None, 65, 1024)          2048      \n",
            " LayerNormalization)                                             \n",
            "                                                                 \n",
            " ExtractToken (Lambda)       (None, 1024)              0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 305525760 (1.14 GB)\n",
            "Trainable params: 305525760 (1.14 GB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for layer in base_model.layers:\n",
        "    layer.trainable=False\n",
        "\n",
        "print(\"Base model layers are freezed!!!\")"
      ],
      "metadata": {
        "id": "5V8VwdtPtaJl",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9f938da3-afce-46c4-db43-7e86c661ac6f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Base model layers are freezed!!!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(base_model.summary())"
      ],
      "metadata": {
        "id": "_6T8gi88tfwQ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "170df96c-ede4-42b3-df1d-14507d068239"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"vit-l32\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 256, 256, 3)]     0         \n",
            "                                                                 \n",
            " embedding (Conv2D)          (None, 8, 8, 1024)        3146752   \n",
            "                                                                 \n",
            " reshape (Reshape)           (None, 64, 1024)          0         \n",
            "                                                                 \n",
            " class_token (ClassToken)    (None, 65, 1024)          1024      \n",
            "                                                                 \n",
            " Transformer/posembed_input  (None, 65, 1024)          66560     \n",
            "  (AddPositionEmbs)                                              \n",
            "                                                                 \n",
            " Transformer/encoderblock_0  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_3  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_4  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_5  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_6  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_7  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_8  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_9  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 0 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 1 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 2 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 3 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 4 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 5 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 6 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 7 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 8 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 9 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 0 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 1 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 2 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 3 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoder_norm (  (None, 65, 1024)          2048      \n",
            " LayerNormalization)                                             \n",
            "                                                                 \n",
            " ExtractToken (Lambda)       (None, 1024)              0         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 305525760 (1.14 GB)\n",
            "Trainable params: 0 (0.00 Byte)\n",
            "Non-trainable params: 305525760 (1.14 GB)\n",
            "_________________________________________________________________\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x = Dropout(0.2)(base_model.output)\n",
        "x = Flatten()(x)\n",
        "x = Dense(2,'softmax')(x)\n",
        "\n",
        "model = Model(inputs=base_model.input,outputs=x)"
      ],
      "metadata": {
        "id": "B15m6TNcthaQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"Final Model is:\\n\",model.summary())"
      ],
      "metadata": {
        "id": "BoXtkAQZt2t0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "30721fea-d5a2-4efd-cc94-401aac261c59"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, 256, 256, 3)]     0         \n",
            "                                                                 \n",
            " embedding (Conv2D)          (None, 8, 8, 1024)        3146752   \n",
            "                                                                 \n",
            " reshape (Reshape)           (None, 64, 1024)          0         \n",
            "                                                                 \n",
            " class_token (ClassToken)    (None, 65, 1024)          1024      \n",
            "                                                                 \n",
            " Transformer/posembed_input  (None, 65, 1024)          66560     \n",
            "  (AddPositionEmbs)                                              \n",
            "                                                                 \n",
            " Transformer/encoderblock_0  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_3  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_4  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_5  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_6  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_7  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_8  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_9  ((None, 65, 1024),        12596224  \n",
            "  (TransformerBlock)          (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 0 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 1 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 2 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 3 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 4 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 5 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 6 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 7 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 8 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_1  ((None, 65, 1024),        12596224  \n",
            " 9 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 0 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 1 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 2 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoderblock_2  ((None, 65, 1024),        12596224  \n",
            " 3 (TransformerBlock)         (None, 16, None, None)             \n",
            "                             )                                   \n",
            "                                                                 \n",
            " Transformer/encoder_norm (  (None, 65, 1024)          2048      \n",
            " LayerNormalization)                                             \n",
            "                                                                 \n",
            " ExtractToken (Lambda)       (None, 1024)              0         \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 1024)              0         \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 1024)              0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 2)                 2050      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 305527810 (1.14 GB)\n",
            "Trainable params: 2050 (8.01 KB)\n",
            "Non-trainable params: 305525760 (1.14 GB)\n",
            "_________________________________________________________________\n",
            "Final Model is:\n",
            " None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "opt=Adam(learning_rate=0.001, beta_1=0.9, beta_2=0.999, decay=0.0, amsgrad=False)"
      ],
      "metadata": {
        "id": "urIYHcjmt4I9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy',optimizer=opt,metrics=['accuracy'])"
      ],
      "metadata": {
        "id": "sFjmGzb5t7Ba"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "RANDOM_SEED=123\n",
        "train_datagen = ImageDataGenerator(\n",
        "    width_shift_range=0.4,\n",
        "    height_shift_range=0.4,\n",
        "    rescale=1.0/255,\n",
        "    horizontal_flip=True\n",
        ")\n",
        "\n",
        "test_datagen = ImageDataGenerator(\n",
        "    rescale=1.0/255,\n",
        ")\n",
        "\n",
        "\n",
        "train_generator = train_datagen.flow_from_directory(\n",
        "    train_dir,\n",
        "    color_mode='rgb',\n",
        "    target_size=img_size,\n",
        "    batch_size=64,\n",
        "    class_mode='categorical',\n",
        "    seed=RANDOM_SEED\n",
        ")\n",
        "\n",
        "\n",
        "validation_generator = test_datagen.flow_from_directory(\n",
        "    test_dir,\n",
        "    color_mode='rgb',\n",
        "    target_size=img_size,\n",
        "    batch_size=64,\n",
        "    class_mode='categorical',\n",
        "    seed=RANDOM_SEED\n",
        ")"
      ],
      "metadata": {
        "id": "tUe3RvE8t_vw",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "422b007b-2d98-4b66-b5e6-b09aa79e9dca"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 20000 images belonging to 2 classes.\n",
            "Found 5000 images belonging to 2 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**3) Train Model**"
      ],
      "metadata": {
        "id": "fElJaSQbvkeE"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import time\n",
        "start = time.time()\n",
        "hist = model.fit(\n",
        "    train_generator,\n",
        "    epochs=15,\n",
        "    validation_data = validation_generator\n",
        ")\n",
        "\n",
        "end = time.time()\n",
        "print(\"Total time taken: \",end-start)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Wt1Jer8UuBKV",
        "outputId": "4abb5003-dbbb-4ff4-c970-de121bdb3b29"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/15\n",
            "313/313 [==============================] - 3646s 12s/step - loss: 0.0854 - accuracy: 0.9669 - val_loss: 0.0264 - val_accuracy: 0.9904\n",
            "Epoch 2/15\n",
            "313/313 [==============================] - 433s 1s/step - loss: 0.0844 - accuracy: 0.9678 - val_loss: 0.0282 - val_accuracy: 0.9906\n",
            "Epoch 3/15\n",
            "313/313 [==============================] - 436s 1s/step - loss: 0.0779 - accuracy: 0.9691 - val_loss: 0.0274 - val_accuracy: 0.9910\n",
            "Epoch 4/15\n",
            "313/313 [==============================] - 438s 1s/step - loss: 0.0732 - accuracy: 0.9718 - val_loss: 0.0278 - val_accuracy: 0.9910\n",
            "Epoch 5/15\n",
            "313/313 [==============================] - 438s 1s/step - loss: 0.0803 - accuracy: 0.9693 - val_loss: 0.0278 - val_accuracy: 0.9900\n",
            "Epoch 6/15\n",
            "313/313 [==============================] - 435s 1s/step - loss: 0.0773 - accuracy: 0.9700 - val_loss: 0.0257 - val_accuracy: 0.9916\n",
            "Epoch 7/15\n",
            "313/313 [==============================] - 427s 1s/step - loss: 0.0780 - accuracy: 0.9701 - val_loss: 0.0269 - val_accuracy: 0.9902\n",
            "Epoch 8/15\n",
            "313/313 [==============================] - 429s 1s/step - loss: 0.0764 - accuracy: 0.9717 - val_loss: 0.0250 - val_accuracy: 0.9918\n",
            "Epoch 9/15\n",
            "313/313 [==============================] - 433s 1s/step - loss: 0.0780 - accuracy: 0.9704 - val_loss: 0.0253 - val_accuracy: 0.9904\n",
            "Epoch 10/15\n",
            "313/313 [==============================] - 426s 1s/step - loss: 0.0812 - accuracy: 0.9689 - val_loss: 0.0258 - val_accuracy: 0.9910\n",
            "Epoch 11/15\n",
            "313/313 [==============================] - 428s 1s/step - loss: 0.0787 - accuracy: 0.9693 - val_loss: 0.0271 - val_accuracy: 0.9898\n",
            "Epoch 12/15\n",
            "313/313 [==============================] - 428s 1s/step - loss: 0.0812 - accuracy: 0.9674 - val_loss: 0.0262 - val_accuracy: 0.9906\n",
            "Epoch 13/15\n",
            "313/313 [==============================] - 438s 1s/step - loss: 0.0805 - accuracy: 0.9692 - val_loss: 0.0272 - val_accuracy: 0.9892\n",
            "Epoch 14/15\n",
            "313/313 [==============================] - 430s 1s/step - loss: 0.0774 - accuracy: 0.9702 - val_loss: 0.0281 - val_accuracy: 0.9884\n",
            "Epoch 15/15\n",
            "313/313 [==============================] - 430s 1s/step - loss: 0.0779 - accuracy: 0.9701 - val_loss: 0.0255 - val_accuracy: 0.9892\n",
            "Total time taken:  9697.201612234116\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**4) Plot Graph**"
      ],
      "metadata": {
        "id": "CS5nRG55vq0X"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# plotting the figure for accuracy\n",
        "plt.figure(figsize=(10,7))\n",
        "plt.plot(hist.history['accuracy'])\n",
        "plt.plot(hist.history['val_accuracy'])\n",
        "plt.title('Model Accuracy Vision Transformer')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='lower right')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "9hsFC3DRuFdV",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 641
        },
        "outputId": "2ae64790-6382-45b6-ce76-0d431bdb7e1c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# plotting the figure for loss\n",
        "plt.figure(figsize=(10,7))\n",
        "plt.plot(hist.history['loss'])\n",
        "plt.plot(hist.history['val_loss'])\n",
        "plt.title('Model Loss Vision Transformer')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "gKE-YMTvuIHP",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 641
        },
        "outputId": "7ba96fa6-1cb9-45fa-cc20-cd6c27361e99"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.save('dogCat_vit_l32.h5')"
      ],
      "metadata": {
        "id": "_f5ezsjduJXY",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "da4efa4f-ac89-400f-d440-d186e8df3279"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        }
      ]
    }
  ]
}