image_classification_ViT.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Opd4X7uzyaEZ"
      },
      "source": [
        "# Image classification with Vision Transformer"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JHLlfvhmyaEd"
      },
      "source": [
        "## Introduction\n",
        "\n",
        "This example implements the [Vision Transformer (ViT)](https://arxiv.org/abs/2010.11929)\n",
        "model by Alexey Dosovitskiy et al. for image classification,\n",
        "and demonstrates it on the CIFAR-100 dataset.\n",
        "The ViT model applies the Transformer architecture with self-attention to sequences of\n",
        "image patches, without using convolution layers."
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "![image.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "9wHn7-_9cqRH"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rHXeznwcyaEe"
      },
      "source": [
        "## Setup"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EjqfixlE34_c",
        "outputId": "c00fed01-73a1-47a9-a222-c19f29697207"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: tensorflow-addons in /usr/local/lib/python3.10/dist-packages (0.23.0)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.10/dist-packages (from tensorflow-addons) (24.0)\n",
            "Requirement already satisfied: typeguard<3.0.0,>=2.7 in /usr/local/lib/python3.10/dist-packages (from tensorflow-addons) (2.13.3)\n"
          ]
        }
      ],
      "source": [
        "pip install -U tensorflow-addons"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "7ckbTG1JyaEe",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "aa204545-3422-442f-f9b2-b87576dab2fa"
      },
      "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"
          ]
        }
      ],
      "source": [
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "import tensorflow_addons as tfa"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R_PiaqPKyaEg"
      },
      "source": [
        "## Prepare the data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "484S56nKyaEg",
        "outputId": "66dd6da9-4a18-4217-f9d8-5aebdc453349"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "x_train shape: (50000, 32, 32, 3) - y_train shape: (50000, 1)\n",
            "x_test shape: (10000, 32, 32, 3) - y_test shape: (10000, 1)\n"
          ]
        }
      ],
      "source": [
        "num_classes = 100\n",
        "input_shape = (32, 32, 3)\n",
        "\n",
        "(x_train, y_train), (x_test, y_test) = keras.datasets.cifar100.load_data()\n",
        "\n",
        "print(f\"x_train shape: {x_train.shape} - y_train shape: {y_train.shape}\")\n",
        "print(f\"x_test shape: {x_test.shape} - y_test shape: {y_test.shape}\")\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RoAeu6bnyaEh"
      },
      "source": [
        "## Configure the hyperparameters"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "UWqco5yuyaEh"
      },
      "outputs": [],
      "source": [
        "learning_rate = 0.001\n",
        "weight_decay = 0.0001\n",
        "batch_size = 256\n",
        "num_epochs = 100\n",
        "image_size = 64  # We'll resize input images to this size\n",
        "patch_size = 8  # Size of the patches to be extract from the input images\n",
        "num_patches = (image_size // patch_size) ** 2\n",
        "projection_dim = 64\n",
        "num_heads = 2\n",
        "transformer_units = [\n",
        "    projection_dim * 2,\n",
        "    projection_dim,\n",
        "]  # Size of the transformer layers\n",
        "transformer_layers = 8\n",
        "mlp_head_units = [1024, 512]  # Size of the dense layers of the final classifier\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "- The hyperparameters can be modified depending on the size of the image and number of patches that can be formed from the image.\n",
        "- Transformer layers and MLP head units can be increased if the dataset is huge."
      ],
      "metadata": {
        "id": "jwEARiWudKM8"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NEHkrAqxyaEi"
      },
      "source": [
        "## Use data augmentation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "t-B8qhQCyaEi"
      },
      "outputs": [],
      "source": [
        "data_augmentation = keras.Sequential(\n",
        "    [\n",
        "        layers.Normalization(),\n",
        "        layers.Resizing(image_size, image_size),\n",
        "        layers.RandomFlip(\"horizontal\"),\n",
        "        layers.RandomRotation(factor=0.02),\n",
        "        layers.RandomZoom(\n",
        "            height_factor=0.2, width_factor=0.2\n",
        "        ),\n",
        "    ],\n",
        "    name=\"data_augmentation\",\n",
        ")\n",
        "# Compute the mean and the variance of the training data for normalization.\n",
        "data_augmentation.layers[0].adapt(x_train)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AXmhcx_cyaEj"
      },
      "source": [
        "## Implement multilayer perceptron (MLP)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "gsoxlMG4yaEj"
      },
      "outputs": [],
      "source": [
        "def mlp(x, hidden_units, dropout_rate):\n",
        "    for units in hidden_units:\n",
        "        x = layers.Dense(units, activation=tf.nn.gelu)(x)\n",
        "        x = layers.Dropout(dropout_rate)(x)\n",
        "    return x"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5wCgOG9byaEj"
      },
      "source": [
        "## Implement patch creation as a layer"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "9w-lYbkByaEj"
      },
      "outputs": [],
      "source": [
        "class Patches(layers.Layer):\n",
        "    def __init__(self, patch_size):\n",
        "        super(Patches, self).__init__()\n",
        "        self.patch_size = patch_size\n",
        "\n",
        "    def call(self, images):\n",
        "        batch_size = tf.shape(images)[0]\n",
        "        patches = tf.image.extract_patches(\n",
        "            images=images,\n",
        "            sizes=[1, self.patch_size, self.patch_size, 1],\n",
        "            strides=[1, self.patch_size, self.patch_size, 1],\n",
        "            rates=[1, 1, 1, 1],\n",
        "            padding=\"VALID\",\n",
        "        )\n",
        "        patch_dims = patches.shape[-1]\n",
        "        patches = tf.reshape(patches, [batch_size, -1, patch_dims])\n",
        "        return patches"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "u4fHthVSyaEj"
      },
      "source": [
        "Let's display patches for a sample image"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 742
        },
        "id": "unMWuumJyaEk",
        "outputId": "323ffdb7-68df-41c9-d159-d6f4a5daec09"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Image size: 64 X 64\n",
            "Patch size: 8 X 8\n",
            "Patches per image: 64\n",
            "Elements per patch: 192\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 400x400 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 400x400 with 64 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.figure(figsize=(4, 4))\n",
        "image = x_train[np.random.choice(range(x_train.shape[0]))]\n",
        "plt.imshow(image.astype(\"uint8\"))\n",
        "plt.axis(\"off\")\n",
        "\n",
        "resized_image = tf.image.resize(\n",
        "    tf.convert_to_tensor([image]), size=(image_size, image_size)\n",
        ")\n",
        "patches = Patches(patch_size)(resized_image)\n",
        "print(f\"Image size: {image_size} X {image_size}\")\n",
        "print(f\"Patch size: {patch_size} X {patch_size}\")\n",
        "print(f\"Patches per image: {patches.shape[1]}\")\n",
        "print(f\"Elements per patch: {patches.shape[-1]}\")\n",
        "\n",
        "n = int(np.sqrt(patches.shape[1]))\n",
        "plt.figure(figsize=(4, 4))\n",
        "for i, patch in enumerate(patches[0]):\n",
        "    ax = plt.subplot(n, n, i + 1)\n",
        "    patch_img = tf.reshape(patch, (patch_size, patch_size, 3))\n",
        "    plt.imshow(patch_img.numpy().astype(\"uint8\"))\n",
        "    plt.axis(\"off\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jRafwW8xyaEk"
      },
      "source": [
        "## Implement the patch encoding layer\n",
        "\n",
        "The `PatchEncoder` layer will linearly transform a patch by projecting it into a\n",
        "vector of size `projection_dim`. In addition, it adds a learnable position\n",
        "embedding to the projected vector."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "Q34kN-xsyaEk"
      },
      "outputs": [],
      "source": [
        "class PatchEncoder(layers.Layer):\n",
        "    def __init__(self, num_patches, projection_dim):\n",
        "        super(PatchEncoder, self).__init__()\n",
        "        self.num_patches = num_patches\n",
        "        self.projection = layers.Dense(units=projection_dim)\n",
        "        self.position_embedding = layers.Embedding(\n",
        "            input_dim=num_patches, output_dim=projection_dim\n",
        "        )\n",
        "\n",
        "    def call(self, patch):\n",
        "        positions = tf.range(start=0, limit=self.num_patches, delta=1)\n",
        "        encoded = self.projection(patch) + self.position_embedding(positions)\n",
        "        return encoded"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IUeBguYByaEk"
      },
      "source": [
        "## Build the ViT model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "vMxk7f2hyaEl"
      },
      "outputs": [],
      "source": [
        "def create_vit_classifier():\n",
        "    inputs = layers.Input(shape=input_shape)\n",
        "    # Augment data.\n",
        "    augmented = data_augmentation(inputs)\n",
        "    # Create patches.\n",
        "    patches = Patches(patch_size)(augmented)\n",
        "    # Encode patches.\n",
        "    encoded_patches = PatchEncoder(num_patches, projection_dim)(patches)\n",
        "\n",
        "    # Create multiple layers of the Transformer block.\n",
        "    for _ in range(transformer_layers):\n",
        "        # Layer normalization 1.\n",
        "        x1 = layers.LayerNormalization(epsilon=1e-6)(encoded_patches)\n",
        "        # Create a multi-head attention layer.\n",
        "        attention_output = layers.MultiHeadAttention(\n",
        "            num_heads=num_heads, key_dim=projection_dim, dropout=0.2\n",
        "        )(x1, x1)\n",
        "        # Skip connection 1.\n",
        "        x2 = layers.Add()([attention_output, encoded_patches])\n",
        "        # Layer normalization 2.\n",
        "        x3 = layers.LayerNormalization(epsilon=1e-6)(x2)\n",
        "        # MLP.\n",
        "        x3 = mlp(x3, hidden_units=transformer_units, dropout_rate=0.1)\n",
        "        # Skip connection 2.\n",
        "        encoded_patches = layers.Add()([x3, x2])\n",
        "\n",
        "    # Create a [batch_size, projection_dim] tensor.\n",
        "    representation = layers.LayerNormalization(epsilon=1e-6)(encoded_patches)\n",
        "    representation = layers.Flatten()(representation)\n",
        "    representation = layers.Dropout(0.5)(representation)\n",
        "    # Add MLP.\n",
        "    features = mlp(representation, hidden_units=mlp_head_units, dropout_rate=0.4)\n",
        "    # Classify outputs.\n",
        "    logits = layers.Dense(num_classes)(features)\n",
        "    # Create the Keras model.\n",
        "    model = keras.Model(inputs=inputs, outputs=logits)\n",
        "    return model"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "create_vit_classifier().summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sfrBClwQOE5s",
        "outputId": "e85e98c3-672a-4ea5-f3ab-6744a60d8fef"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                Output Shape                 Param #   Connected to                  \n",
            "==================================================================================================\n",
            " input_1 (InputLayer)        [(None, 32, 32, 3)]          0         []                            \n",
            "                                                                                                  \n",
            " data_augmentation (Sequent  (None, 64, 64, 3)            7         ['input_1[0][0]']             \n",
            " ial)                                                                                             \n",
            "                                                                                                  \n",
            " patches_1 (Patches)         (None, None, 192)            0         ['data_augmentation[0][0]']   \n",
            "                                                                                                  \n",
            " patch_encoder (PatchEncode  (None, 64, 64)               16448     ['patches_1[0][0]']           \n",
            " r)                                                                                               \n",
            "                                                                                                  \n",
            " layer_normalization (Layer  (None, 64, 64)               128       ['patch_encoder[0][0]']       \n",
            " Normalization)                                                                                   \n",
            "                                                                                                  \n",
            " multi_head_attention (Mult  (None, 64, 64)               33216     ['layer_normalization[0][0]', \n",
            " iHeadAttention)                                                     'layer_normalization[0][0]'] \n",
            "                                                                                                  \n",
            " add (Add)                   (None, 64, 64)               0         ['multi_head_attention[0][0]',\n",
            "                                                                     'patch_encoder[0][0]']       \n",
            "                                                                                                  \n",
            " layer_normalization_1 (Lay  (None, 64, 64)               128       ['add[0][0]']                 \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " dense_1 (Dense)             (None, 64, 128)              8320      ['layer_normalization_1[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " dropout (Dropout)           (None, 64, 128)              0         ['dense_1[0][0]']             \n",
            "                                                                                                  \n",
            " dense_2 (Dense)             (None, 64, 64)               8256      ['dropout[0][0]']             \n",
            "                                                                                                  \n",
            " dropout_1 (Dropout)         (None, 64, 64)               0         ['dense_2[0][0]']             \n",
            "                                                                                                  \n",
            " add_1 (Add)                 (None, 64, 64)               0         ['dropout_1[0][0]',           \n",
            "                                                                     'add[0][0]']                 \n",
            "                                                                                                  \n",
            " layer_normalization_2 (Lay  (None, 64, 64)               128       ['add_1[0][0]']               \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " multi_head_attention_1 (Mu  (None, 64, 64)               33216     ['layer_normalization_2[0][0]'\n",
            " ltiHeadAttention)                                                  , 'layer_normalization_2[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " add_2 (Add)                 (None, 64, 64)               0         ['multi_head_attention_1[0][0]\n",
            "                                                                    ',                            \n",
            "                                                                     'add_1[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_3 (Lay  (None, 64, 64)               128       ['add_2[0][0]']               \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " dense_3 (Dense)             (None, 64, 128)              8320      ['layer_normalization_3[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " dropout_2 (Dropout)         (None, 64, 128)              0         ['dense_3[0][0]']             \n",
            "                                                                                                  \n",
            " dense_4 (Dense)             (None, 64, 64)               8256      ['dropout_2[0][0]']           \n",
            "                                                                                                  \n",
            " dropout_3 (Dropout)         (None, 64, 64)               0         ['dense_4[0][0]']             \n",
            "                                                                                                  \n",
            " add_3 (Add)                 (None, 64, 64)               0         ['dropout_3[0][0]',           \n",
            "                                                                     'add_2[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_4 (Lay  (None, 64, 64)               128       ['add_3[0][0]']               \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " multi_head_attention_2 (Mu  (None, 64, 64)               33216     ['layer_normalization_4[0][0]'\n",
            " ltiHeadAttention)                                                  , 'layer_normalization_4[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " add_4 (Add)                 (None, 64, 64)               0         ['multi_head_attention_2[0][0]\n",
            "                                                                    ',                            \n",
            "                                                                     'add_3[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_5 (Lay  (None, 64, 64)               128       ['add_4[0][0]']               \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " dense_5 (Dense)             (None, 64, 128)              8320      ['layer_normalization_5[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " dropout_4 (Dropout)         (None, 64, 128)              0         ['dense_5[0][0]']             \n",
            "                                                                                                  \n",
            " dense_6 (Dense)             (None, 64, 64)               8256      ['dropout_4[0][0]']           \n",
            "                                                                                                  \n",
            " dropout_5 (Dropout)         (None, 64, 64)               0         ['dense_6[0][0]']             \n",
            "                                                                                                  \n",
            " add_5 (Add)                 (None, 64, 64)               0         ['dropout_5[0][0]',           \n",
            "                                                                     'add_4[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_6 (Lay  (None, 64, 64)               128       ['add_5[0][0]']               \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " multi_head_attention_3 (Mu  (None, 64, 64)               33216     ['layer_normalization_6[0][0]'\n",
            " ltiHeadAttention)                                                  , 'layer_normalization_6[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " add_6 (Add)                 (None, 64, 64)               0         ['multi_head_attention_3[0][0]\n",
            "                                                                    ',                            \n",
            "                                                                     'add_5[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_7 (Lay  (None, 64, 64)               128       ['add_6[0][0]']               \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " dense_7 (Dense)             (None, 64, 128)              8320      ['layer_normalization_7[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " dropout_6 (Dropout)         (None, 64, 128)              0         ['dense_7[0][0]']             \n",
            "                                                                                                  \n",
            " dense_8 (Dense)             (None, 64, 64)               8256      ['dropout_6[0][0]']           \n",
            "                                                                                                  \n",
            " dropout_7 (Dropout)         (None, 64, 64)               0         ['dense_8[0][0]']             \n",
            "                                                                                                  \n",
            " add_7 (Add)                 (None, 64, 64)               0         ['dropout_7[0][0]',           \n",
            "                                                                     'add_6[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_8 (Lay  (None, 64, 64)               128       ['add_7[0][0]']               \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " multi_head_attention_4 (Mu  (None, 64, 64)               33216     ['layer_normalization_8[0][0]'\n",
            " ltiHeadAttention)                                                  , 'layer_normalization_8[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " add_8 (Add)                 (None, 64, 64)               0         ['multi_head_attention_4[0][0]\n",
            "                                                                    ',                            \n",
            "                                                                     'add_7[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_9 (Lay  (None, 64, 64)               128       ['add_8[0][0]']               \n",
            " erNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " dense_9 (Dense)             (None, 64, 128)              8320      ['layer_normalization_9[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " dropout_8 (Dropout)         (None, 64, 128)              0         ['dense_9[0][0]']             \n",
            "                                                                                                  \n",
            " dense_10 (Dense)            (None, 64, 64)               8256      ['dropout_8[0][0]']           \n",
            "                                                                                                  \n",
            " dropout_9 (Dropout)         (None, 64, 64)               0         ['dense_10[0][0]']            \n",
            "                                                                                                  \n",
            " add_9 (Add)                 (None, 64, 64)               0         ['dropout_9[0][0]',           \n",
            "                                                                     'add_8[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_10 (La  (None, 64, 64)               128       ['add_9[0][0]']               \n",
            " yerNormalization)                                                                                \n",
            "                                                                                                  \n",
            " multi_head_attention_5 (Mu  (None, 64, 64)               33216     ['layer_normalization_10[0][0]\n",
            " ltiHeadAttention)                                                  ',                            \n",
            "                                                                     'layer_normalization_10[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " add_10 (Add)                (None, 64, 64)               0         ['multi_head_attention_5[0][0]\n",
            "                                                                    ',                            \n",
            "                                                                     'add_9[0][0]']               \n",
            "                                                                                                  \n",
            " layer_normalization_11 (La  (None, 64, 64)               128       ['add_10[0][0]']              \n",
            " yerNormalization)                                                                                \n",
            "                                                                                                  \n",
            " dense_11 (Dense)            (None, 64, 128)              8320      ['layer_normalization_11[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " dropout_10 (Dropout)        (None, 64, 128)              0         ['dense_11[0][0]']            \n",
            "                                                                                                  \n",
            " dense_12 (Dense)            (None, 64, 64)               8256      ['dropout_10[0][0]']          \n",
            "                                                                                                  \n",
            " dropout_11 (Dropout)        (None, 64, 64)               0         ['dense_12[0][0]']            \n",
            "                                                                                                  \n",
            " add_11 (Add)                (None, 64, 64)               0         ['dropout_11[0][0]',          \n",
            "                                                                     'add_10[0][0]']              \n",
            "                                                                                                  \n",
            " layer_normalization_12 (La  (None, 64, 64)               128       ['add_11[0][0]']              \n",
            " yerNormalization)                                                                                \n",
            "                                                                                                  \n",
            " multi_head_attention_6 (Mu  (None, 64, 64)               33216     ['layer_normalization_12[0][0]\n",
            " ltiHeadAttention)                                                  ',                            \n",
            "                                                                     'layer_normalization_12[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " add_12 (Add)                (None, 64, 64)               0         ['multi_head_attention_6[0][0]\n",
            "                                                                    ',                            \n",
            "                                                                     'add_11[0][0]']              \n",
            "                                                                                                  \n",
            " layer_normalization_13 (La  (None, 64, 64)               128       ['add_12[0][0]']              \n",
            " yerNormalization)                                                                                \n",
            "                                                                                                  \n",
            " dense_13 (Dense)            (None, 64, 128)              8320      ['layer_normalization_13[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " dropout_12 (Dropout)        (None, 64, 128)              0         ['dense_13[0][0]']            \n",
            "                                                                                                  \n",
            " dense_14 (Dense)            (None, 64, 64)               8256      ['dropout_12[0][0]']          \n",
            "                                                                                                  \n",
            " dropout_13 (Dropout)        (None, 64, 64)               0         ['dense_14[0][0]']            \n",
            "                                                                                                  \n",
            " add_13 (Add)                (None, 64, 64)               0         ['dropout_13[0][0]',          \n",
            "                                                                     'add_12[0][0]']              \n",
            "                                                                                                  \n",
            " layer_normalization_14 (La  (None, 64, 64)               128       ['add_13[0][0]']              \n",
            " yerNormalization)                                                                                \n",
            "                                                                                                  \n",
            " multi_head_attention_7 (Mu  (None, 64, 64)               33216     ['layer_normalization_14[0][0]\n",
            " ltiHeadAttention)                                                  ',                            \n",
            "                                                                     'layer_normalization_14[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " add_14 (Add)                (None, 64, 64)               0         ['multi_head_attention_7[0][0]\n",
            "                                                                    ',                            \n",
            "                                                                     'add_13[0][0]']              \n",
            "                                                                                                  \n",
            " layer_normalization_15 (La  (None, 64, 64)               128       ['add_14[0][0]']              \n",
            " yerNormalization)                                                                                \n",
            "                                                                                                  \n",
            " dense_15 (Dense)            (None, 64, 128)              8320      ['layer_normalization_15[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " dropout_14 (Dropout)        (None, 64, 128)              0         ['dense_15[0][0]']            \n",
            "                                                                                                  \n",
            " dense_16 (Dense)            (None, 64, 64)               8256      ['dropout_14[0][0]']          \n",
            "                                                                                                  \n",
            " dropout_15 (Dropout)        (None, 64, 64)               0         ['dense_16[0][0]']            \n",
            "                                                                                                  \n",
            " add_15 (Add)                (None, 64, 64)               0         ['dropout_15[0][0]',          \n",
            "                                                                     'add_14[0][0]']              \n",
            "                                                                                                  \n",
            " layer_normalization_16 (La  (None, 64, 64)               128       ['add_15[0][0]']              \n",
            " yerNormalization)                                                                                \n",
            "                                                                                                  \n",
            " flatten (Flatten)           (None, 4096)                 0         ['layer_normalization_16[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " dropout_16 (Dropout)        (None, 4096)                 0         ['flatten[0][0]']             \n",
            "                                                                                                  \n",
            " dense_17 (Dense)            (None, 1024)                 4195328   ['dropout_16[0][0]']          \n",
            "                                                                                                  \n",
            " dropout_17 (Dropout)        (None, 1024)                 0         ['dense_17[0][0]']            \n",
            "                                                                                                  \n",
            " dense_18 (Dense)            (None, 512)                  524800    ['dropout_17[0][0]']          \n",
            "                                                                                                  \n",
            " dropout_18 (Dropout)        (None, 512)                  0         ['dense_18[0][0]']            \n",
            "                                                                                                  \n",
            " dense_19 (Dense)            (None, 100)                  51300     ['dropout_18[0][0]']          \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 5188395 (19.79 MB)\n",
            "Trainable params: 5188388 (19.79 MB)\n",
            "Non-trainable params: 7 (32.00 Byte)\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LBFiiiS4yaEl"
      },
      "source": [
        "## Compile, train, and evaluate the mode"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KUG7jajoyaEl",
        "outputId": "c5066416-00be-4363-f28e-fe0576f48fde"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "176/176 [==============================] - 37s 121ms/step - loss: 4.3366 - accuracy: 0.0555 - top-5-accuracy: 0.1852 - val_loss: 3.8081 - val_accuracy: 0.1204 - val_top-5-accuracy: 0.3454\n",
            "Epoch 2/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 3.8254 - accuracy: 0.1141 - top-5-accuracy: 0.3294 - val_loss: 3.4757 - val_accuracy: 0.1774 - val_top-5-accuracy: 0.4446\n",
            "Epoch 3/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 3.5598 - accuracy: 0.1550 - top-5-accuracy: 0.4072 - val_loss: 3.2559 - val_accuracy: 0.2158 - val_top-5-accuracy: 0.4864\n",
            "Epoch 4/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 3.3870 - accuracy: 0.1854 - top-5-accuracy: 0.4573 - val_loss: 3.1018 - val_accuracy: 0.2458 - val_top-5-accuracy: 0.5256\n",
            "Epoch 5/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 3.2528 - accuracy: 0.2126 - top-5-accuracy: 0.4907 - val_loss: 2.9780 - val_accuracy: 0.2642 - val_top-5-accuracy: 0.5570\n",
            "Epoch 6/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 3.1372 - accuracy: 0.2319 - top-5-accuracy: 0.5224 - val_loss: 2.9039 - val_accuracy: 0.2832 - val_top-5-accuracy: 0.5764\n",
            "Epoch 7/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 3.0516 - accuracy: 0.2471 - top-5-accuracy: 0.5438 - val_loss: 2.7782 - val_accuracy: 0.3072 - val_top-5-accuracy: 0.6112\n",
            "Epoch 8/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 2.9575 - accuracy: 0.2679 - top-5-accuracy: 0.5681 - val_loss: 2.7247 - val_accuracy: 0.3242 - val_top-5-accuracy: 0.6150\n",
            "Epoch 9/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 2.8707 - accuracy: 0.2800 - top-5-accuracy: 0.5875 - val_loss: 2.6400 - val_accuracy: 0.3338 - val_top-5-accuracy: 0.6402\n",
            "Epoch 10/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 2.7986 - accuracy: 0.2952 - top-5-accuracy: 0.6034 - val_loss: 2.6104 - val_accuracy: 0.3418 - val_top-5-accuracy: 0.6452\n",
            "Epoch 11/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 2.7101 - accuracy: 0.3112 - top-5-accuracy: 0.6262 - val_loss: 2.5198 - val_accuracy: 0.3562 - val_top-5-accuracy: 0.6626\n",
            "Epoch 12/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 2.6349 - accuracy: 0.3279 - top-5-accuracy: 0.6429 - val_loss: 2.4690 - val_accuracy: 0.3658 - val_top-5-accuracy: 0.6712\n",
            "Epoch 13/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 2.5640 - accuracy: 0.3406 - top-5-accuracy: 0.6576 - val_loss: 2.3961 - val_accuracy: 0.3856 - val_top-5-accuracy: 0.6842\n",
            "Epoch 14/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 2.4896 - accuracy: 0.3563 - top-5-accuracy: 0.6725 - val_loss: 2.3338 - val_accuracy: 0.3986 - val_top-5-accuracy: 0.6994\n",
            "Epoch 15/100\n",
            "176/176 [==============================] - 20s 111ms/step - loss: 2.4321 - accuracy: 0.3709 - top-5-accuracy: 0.6864 - val_loss: 2.3065 - val_accuracy: 0.4020 - val_top-5-accuracy: 0.7110\n",
            "Epoch 16/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 2.3745 - accuracy: 0.3823 - top-5-accuracy: 0.6964 - val_loss: 2.2604 - val_accuracy: 0.4190 - val_top-5-accuracy: 0.7124\n",
            "Epoch 17/100\n",
            "176/176 [==============================] - 20s 114ms/step - loss: 2.3238 - accuracy: 0.3911 - top-5-accuracy: 0.7092 - val_loss: 2.2157 - val_accuracy: 0.4224 - val_top-5-accuracy: 0.7300\n",
            "Epoch 18/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 2.2582 - accuracy: 0.4058 - top-5-accuracy: 0.7228 - val_loss: 2.1695 - val_accuracy: 0.4330 - val_top-5-accuracy: 0.7384\n",
            "Epoch 19/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 2.2243 - accuracy: 0.4119 - top-5-accuracy: 0.7288 - val_loss: 2.1437 - val_accuracy: 0.4390 - val_top-5-accuracy: 0.7378\n",
            "Epoch 20/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 2.1859 - accuracy: 0.4207 - top-5-accuracy: 0.7386 - val_loss: 2.1527 - val_accuracy: 0.4346 - val_top-5-accuracy: 0.7338\n",
            "Epoch 21/100\n",
            "176/176 [==============================] - 19s 108ms/step - loss: 2.1508 - accuracy: 0.4287 - top-5-accuracy: 0.7440 - val_loss: 2.1093 - val_accuracy: 0.4476 - val_top-5-accuracy: 0.7428\n",
            "Epoch 22/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 2.0998 - accuracy: 0.4369 - top-5-accuracy: 0.7571 - val_loss: 2.1083 - val_accuracy: 0.4452 - val_top-5-accuracy: 0.7464\n",
            "Epoch 23/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 2.0685 - accuracy: 0.4437 - top-5-accuracy: 0.7606 - val_loss: 2.0674 - val_accuracy: 0.4532 - val_top-5-accuracy: 0.7558\n",
            "Epoch 24/100\n",
            "176/176 [==============================] - 19s 106ms/step - loss: 2.0262 - accuracy: 0.4542 - top-5-accuracy: 0.7693 - val_loss: 2.0896 - val_accuracy: 0.4470 - val_top-5-accuracy: 0.7542\n",
            "Epoch 25/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 1.9994 - accuracy: 0.4618 - top-5-accuracy: 0.7727 - val_loss: 2.0426 - val_accuracy: 0.4622 - val_top-5-accuracy: 0.7600\n",
            "Epoch 26/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.9745 - accuracy: 0.4652 - top-5-accuracy: 0.7800 - val_loss: 2.0249 - val_accuracy: 0.4646 - val_top-5-accuracy: 0.7654\n",
            "Epoch 27/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.9359 - accuracy: 0.4714 - top-5-accuracy: 0.7855 - val_loss: 1.9934 - val_accuracy: 0.4704 - val_top-5-accuracy: 0.7700\n",
            "Epoch 28/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.9159 - accuracy: 0.4829 - top-5-accuracy: 0.7914 - val_loss: 2.0120 - val_accuracy: 0.4692 - val_top-5-accuracy: 0.7626\n",
            "Epoch 29/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.8776 - accuracy: 0.4853 - top-5-accuracy: 0.7991 - val_loss: 1.9670 - val_accuracy: 0.4794 - val_top-5-accuracy: 0.7754\n",
            "Epoch 30/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.8604 - accuracy: 0.4925 - top-5-accuracy: 0.8019 - val_loss: 1.9421 - val_accuracy: 0.4808 - val_top-5-accuracy: 0.7780\n",
            "Epoch 31/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.8203 - accuracy: 0.4996 - top-5-accuracy: 0.8081 - val_loss: 1.9531 - val_accuracy: 0.4848 - val_top-5-accuracy: 0.7726\n",
            "Epoch 32/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.8128 - accuracy: 0.5018 - top-5-accuracy: 0.8109 - val_loss: 1.9391 - val_accuracy: 0.4840 - val_top-5-accuracy: 0.7812\n",
            "Epoch 33/100\n",
            "176/176 [==============================] - 19s 106ms/step - loss: 1.7904 - accuracy: 0.5066 - top-5-accuracy: 0.8144 - val_loss: 1.9544 - val_accuracy: 0.4764 - val_top-5-accuracy: 0.7756\n",
            "Epoch 34/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.7617 - accuracy: 0.5116 - top-5-accuracy: 0.8203 - val_loss: 1.9421 - val_accuracy: 0.4908 - val_top-5-accuracy: 0.7762\n",
            "Epoch 35/100\n",
            "176/176 [==============================] - 20s 114ms/step - loss: 1.7447 - accuracy: 0.5168 - top-5-accuracy: 0.8232 - val_loss: 1.9003 - val_accuracy: 0.4938 - val_top-5-accuracy: 0.7882\n",
            "Epoch 36/100\n",
            "176/176 [==============================] - 19s 108ms/step - loss: 1.7237 - accuracy: 0.5210 - top-5-accuracy: 0.8260 - val_loss: 1.8971 - val_accuracy: 0.4956 - val_top-5-accuracy: 0.7916\n",
            "Epoch 37/100\n",
            "176/176 [==============================] - 19s 111ms/step - loss: 1.7048 - accuracy: 0.5252 - top-5-accuracy: 0.8309 - val_loss: 1.9156 - val_accuracy: 0.4978 - val_top-5-accuracy: 0.7864\n",
            "Epoch 38/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.6816 - accuracy: 0.5331 - top-5-accuracy: 0.8346 - val_loss: 1.9038 - val_accuracy: 0.4962 - val_top-5-accuracy: 0.7870\n",
            "Epoch 39/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.6638 - accuracy: 0.5328 - top-5-accuracy: 0.8370 - val_loss: 1.8986 - val_accuracy: 0.4928 - val_top-5-accuracy: 0.7896\n",
            "Epoch 40/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.6473 - accuracy: 0.5390 - top-5-accuracy: 0.8391 - val_loss: 1.8796 - val_accuracy: 0.4978 - val_top-5-accuracy: 0.7880\n",
            "Epoch 41/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.6359 - accuracy: 0.5426 - top-5-accuracy: 0.8429 - val_loss: 1.8692 - val_accuracy: 0.5006 - val_top-5-accuracy: 0.7898\n",
            "Epoch 42/100\n",
            "176/176 [==============================] - 19s 108ms/step - loss: 1.6132 - accuracy: 0.5488 - top-5-accuracy: 0.8462 - val_loss: 1.8642 - val_accuracy: 0.5070 - val_top-5-accuracy: 0.7922\n",
            "Epoch 43/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.5936 - accuracy: 0.5520 - top-5-accuracy: 0.8514 - val_loss: 1.8832 - val_accuracy: 0.5002 - val_top-5-accuracy: 0.7954\n",
            "Epoch 44/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.5763 - accuracy: 0.5556 - top-5-accuracy: 0.8536 - val_loss: 1.8560 - val_accuracy: 0.5092 - val_top-5-accuracy: 0.7920\n",
            "Epoch 45/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.5511 - accuracy: 0.5625 - top-5-accuracy: 0.8559 - val_loss: 1.8551 - val_accuracy: 0.5114 - val_top-5-accuracy: 0.7960\n",
            "Epoch 46/100\n",
            "176/176 [==============================] - 19s 106ms/step - loss: 1.5431 - accuracy: 0.5608 - top-5-accuracy: 0.8596 - val_loss: 1.8558 - val_accuracy: 0.5068 - val_top-5-accuracy: 0.7958\n",
            "Epoch 47/100\n",
            "176/176 [==============================] - 19s 108ms/step - loss: 1.5231 - accuracy: 0.5696 - top-5-accuracy: 0.8607 - val_loss: 1.8560 - val_accuracy: 0.5086 - val_top-5-accuracy: 0.7954\n",
            "Epoch 48/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.5204 - accuracy: 0.5679 - top-5-accuracy: 0.8610 - val_loss: 1.8703 - val_accuracy: 0.5058 - val_top-5-accuracy: 0.7902\n",
            "Epoch 49/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 1.4900 - accuracy: 0.5736 - top-5-accuracy: 0.8676 - val_loss: 1.8335 - val_accuracy: 0.5130 - val_top-5-accuracy: 0.8040\n",
            "Epoch 50/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.4947 - accuracy: 0.5714 - top-5-accuracy: 0.8677 - val_loss: 1.8300 - val_accuracy: 0.5172 - val_top-5-accuracy: 0.8056\n",
            "Epoch 51/100\n",
            "176/176 [==============================] - 19s 108ms/step - loss: 1.4753 - accuracy: 0.5811 - top-5-accuracy: 0.8688 - val_loss: 1.8389 - val_accuracy: 0.5142 - val_top-5-accuracy: 0.8010\n",
            "Epoch 52/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 1.4683 - accuracy: 0.5828 - top-5-accuracy: 0.8694 - val_loss: 1.8219 - val_accuracy: 0.5184 - val_top-5-accuracy: 0.8050\n",
            "Epoch 53/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.4498 - accuracy: 0.5852 - top-5-accuracy: 0.8730 - val_loss: 1.8053 - val_accuracy: 0.5192 - val_top-5-accuracy: 0.8136\n",
            "Epoch 54/100\n",
            "176/176 [==============================] - 20s 114ms/step - loss: 1.4299 - accuracy: 0.5899 - top-5-accuracy: 0.8774 - val_loss: 1.8133 - val_accuracy: 0.5222 - val_top-5-accuracy: 0.8006\n",
            "Epoch 55/100\n",
            "176/176 [==============================] - 20s 114ms/step - loss: 1.4265 - accuracy: 0.5908 - top-5-accuracy: 0.8799 - val_loss: 1.7963 - val_accuracy: 0.5230 - val_top-5-accuracy: 0.8072\n",
            "Epoch 56/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 1.4006 - accuracy: 0.5957 - top-5-accuracy: 0.8816 - val_loss: 1.7791 - val_accuracy: 0.5324 - val_top-5-accuracy: 0.8120\n",
            "Epoch 57/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.4024 - accuracy: 0.5966 - top-5-accuracy: 0.8812 - val_loss: 1.7977 - val_accuracy: 0.5236 - val_top-5-accuracy: 0.8138\n",
            "Epoch 58/100\n",
            "176/176 [==============================] - 19s 107ms/step - loss: 1.3989 - accuracy: 0.5958 - top-5-accuracy: 0.8830 - val_loss: 1.8191 - val_accuracy: 0.5210 - val_top-5-accuracy: 0.8078\n",
            "Epoch 59/100\n",
            "176/176 [==============================] - 19s 108ms/step - loss: 1.3888 - accuracy: 0.5976 - top-5-accuracy: 0.8848 - val_loss: 1.7990 - val_accuracy: 0.5234 - val_top-5-accuracy: 0.8094\n",
            "Epoch 60/100\n",
            "176/176 [==============================] - 19s 107ms/step - loss: 1.3684 - accuracy: 0.6045 - top-5-accuracy: 0.8884 - val_loss: 1.7820 - val_accuracy: 0.5298 - val_top-5-accuracy: 0.8100\n",
            "Epoch 61/100\n",
            "176/176 [==============================] - 20s 111ms/step - loss: 1.3572 - accuracy: 0.6076 - top-5-accuracy: 0.8888 - val_loss: 1.7728 - val_accuracy: 0.5330 - val_top-5-accuracy: 0.8170\n",
            "Epoch 62/100\n",
            "176/176 [==============================] - 19s 107ms/step - loss: 1.3556 - accuracy: 0.6114 - top-5-accuracy: 0.8899 - val_loss: 1.7790 - val_accuracy: 0.5238 - val_top-5-accuracy: 0.8148\n",
            "Epoch 63/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.3482 - accuracy: 0.6128 - top-5-accuracy: 0.8896 - val_loss: 1.7876 - val_accuracy: 0.5316 - val_top-5-accuracy: 0.8166\n",
            "Epoch 64/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 1.3258 - accuracy: 0.6154 - top-5-accuracy: 0.8930 - val_loss: 1.7924 - val_accuracy: 0.5334 - val_top-5-accuracy: 0.8144\n",
            "Epoch 65/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.3157 - accuracy: 0.6188 - top-5-accuracy: 0.8953 - val_loss: 1.7950 - val_accuracy: 0.5302 - val_top-5-accuracy: 0.8120\n",
            "Epoch 66/100\n",
            "176/176 [==============================] - 20s 111ms/step - loss: 1.3188 - accuracy: 0.6176 - top-5-accuracy: 0.8956 - val_loss: 1.7947 - val_accuracy: 0.5284 - val_top-5-accuracy: 0.8138\n",
            "Epoch 67/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.3131 - accuracy: 0.6138 - top-5-accuracy: 0.8965 - val_loss: 1.7761 - val_accuracy: 0.5332 - val_top-5-accuracy: 0.8220\n",
            "Epoch 68/100\n",
            "176/176 [==============================] - 20s 114ms/step - loss: 1.3029 - accuracy: 0.6209 - top-5-accuracy: 0.8976 - val_loss: 1.7533 - val_accuracy: 0.5386 - val_top-5-accuracy: 0.8210\n",
            "Epoch 69/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.2896 - accuracy: 0.6250 - top-5-accuracy: 0.8974 - val_loss: 1.7979 - val_accuracy: 0.5262 - val_top-5-accuracy: 0.8122\n",
            "Epoch 70/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 1.2852 - accuracy: 0.6234 - top-5-accuracy: 0.8998 - val_loss: 1.7709 - val_accuracy: 0.5350 - val_top-5-accuracy: 0.8196\n",
            "Epoch 71/100\n",
            "176/176 [==============================] - 20s 115ms/step - loss: 1.2724 - accuracy: 0.6267 - top-5-accuracy: 0.9016 - val_loss: 1.7602 - val_accuracy: 0.5376 - val_top-5-accuracy: 0.8166\n",
            "Epoch 72/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 1.2744 - accuracy: 0.6272 - top-5-accuracy: 0.9019 - val_loss: 1.7598 - val_accuracy: 0.5406 - val_top-5-accuracy: 0.8196\n",
            "Epoch 73/100\n",
            "176/176 [==============================] - 19s 108ms/step - loss: 1.2578 - accuracy: 0.6310 - top-5-accuracy: 0.9053 - val_loss: 1.7446 - val_accuracy: 0.5406 - val_top-5-accuracy: 0.8168\n",
            "Epoch 74/100\n",
            "176/176 [==============================] - 20s 114ms/step - loss: 1.2529 - accuracy: 0.6334 - top-5-accuracy: 0.9048 - val_loss: 1.7604 - val_accuracy: 0.5476 - val_top-5-accuracy: 0.8202\n",
            "Epoch 75/100\n",
            "176/176 [==============================] - 19s 108ms/step - loss: 1.2547 - accuracy: 0.6335 - top-5-accuracy: 0.9052 - val_loss: 1.7564 - val_accuracy: 0.5456 - val_top-5-accuracy: 0.8182\n",
            "Epoch 76/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.2426 - accuracy: 0.6336 - top-5-accuracy: 0.9068 - val_loss: 1.7496 - val_accuracy: 0.5508 - val_top-5-accuracy: 0.8242\n",
            "Epoch 77/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.2361 - accuracy: 0.6392 - top-5-accuracy: 0.9073 - val_loss: 1.7598 - val_accuracy: 0.5484 - val_top-5-accuracy: 0.8206\n",
            "Epoch 78/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.2248 - accuracy: 0.6403 - top-5-accuracy: 0.9087 - val_loss: 1.7744 - val_accuracy: 0.5382 - val_top-5-accuracy: 0.8214\n",
            "Epoch 79/100\n",
            "176/176 [==============================] - 21s 119ms/step - loss: 1.2219 - accuracy: 0.6400 - top-5-accuracy: 0.9099 - val_loss: 1.7957 - val_accuracy: 0.5360 - val_top-5-accuracy: 0.8134\n",
            "Epoch 80/100\n",
            "176/176 [==============================] - 20s 113ms/step - loss: 1.2200 - accuracy: 0.6433 - top-5-accuracy: 0.9085 - val_loss: 1.7560 - val_accuracy: 0.5480 - val_top-5-accuracy: 0.8230\n",
            "Epoch 81/100\n",
            "176/176 [==============================] - 19s 111ms/step - loss: 1.2193 - accuracy: 0.6400 - top-5-accuracy: 0.9090 - val_loss: 1.7509 - val_accuracy: 0.5488 - val_top-5-accuracy: 0.8188\n",
            "Epoch 82/100\n",
            "176/176 [==============================] - 20s 111ms/step - loss: 1.2003 - accuracy: 0.6475 - top-5-accuracy: 0.9123 - val_loss: 1.7605 - val_accuracy: 0.5374 - val_top-5-accuracy: 0.8168\n",
            "Epoch 83/100\n",
            "176/176 [==============================] - 21s 117ms/step - loss: 1.2203 - accuracy: 0.6400 - top-5-accuracy: 0.9110 - val_loss: 1.7395 - val_accuracy: 0.5432 - val_top-5-accuracy: 0.8212\n",
            "Epoch 84/100\n",
            "176/176 [==============================] - 20s 115ms/step - loss: 1.1829 - accuracy: 0.6507 - top-5-accuracy: 0.9174 - val_loss: 1.7500 - val_accuracy: 0.5484 - val_top-5-accuracy: 0.8244\n",
            "Epoch 85/100\n",
            "176/176 [==============================] - 21s 118ms/step - loss: 1.1994 - accuracy: 0.6439 - top-5-accuracy: 0.9128 - val_loss: 1.7570 - val_accuracy: 0.5360 - val_top-5-accuracy: 0.8170\n",
            "Epoch 86/100\n",
            "176/176 [==============================] - 19s 107ms/step - loss: 1.1870 - accuracy: 0.6495 - top-5-accuracy: 0.9149 - val_loss: 1.7769 - val_accuracy: 0.5398 - val_top-5-accuracy: 0.8194\n",
            "Epoch 87/100\n",
            "176/176 [==============================] - 19s 107ms/step - loss: 1.1838 - accuracy: 0.6495 - top-5-accuracy: 0.9150 - val_loss: 1.7607 - val_accuracy: 0.5502 - val_top-5-accuracy: 0.8248\n",
            "Epoch 88/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.1745 - accuracy: 0.6519 - top-5-accuracy: 0.9182 - val_loss: 1.7593 - val_accuracy: 0.5484 - val_top-5-accuracy: 0.8252\n",
            "Epoch 89/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.1681 - accuracy: 0.6554 - top-5-accuracy: 0.9176 - val_loss: 1.7677 - val_accuracy: 0.5396 - val_top-5-accuracy: 0.8196\n",
            "Epoch 90/100\n",
            "176/176 [==============================] - 22s 123ms/step - loss: 1.1705 - accuracy: 0.6532 - top-5-accuracy: 0.9163 - val_loss: 1.7495 - val_accuracy: 0.5510 - val_top-5-accuracy: 0.8228\n",
            "Epoch 91/100\n",
            "176/176 [==============================] - 20s 115ms/step - loss: 1.1668 - accuracy: 0.6536 - top-5-accuracy: 0.9189 - val_loss: 1.7542 - val_accuracy: 0.5494 - val_top-5-accuracy: 0.8252\n",
            "Epoch 92/100\n",
            "176/176 [==============================] - 20s 112ms/step - loss: 1.1607 - accuracy: 0.6569 - top-5-accuracy: 0.9192 - val_loss: 1.7584 - val_accuracy: 0.5456 - val_top-5-accuracy: 0.8218\n",
            "Epoch 93/100\n",
            "176/176 [==============================] - 19s 107ms/step - loss: 1.1570 - accuracy: 0.6567 - top-5-accuracy: 0.9176 - val_loss: 1.7361 - val_accuracy: 0.5502 - val_top-5-accuracy: 0.8262\n",
            "Epoch 94/100\n",
            "176/176 [==============================] - 19s 109ms/step - loss: 1.1502 - accuracy: 0.6620 - top-5-accuracy: 0.9204 - val_loss: 1.7396 - val_accuracy: 0.5506 - val_top-5-accuracy: 0.8270\n",
            "Epoch 95/100\n",
            "176/176 [==============================] - 19s 106ms/step - loss: 1.1489 - accuracy: 0.6578 - top-5-accuracy: 0.9198 - val_loss: 1.7585 - val_accuracy: 0.5468 - val_top-5-accuracy: 0.8210\n",
            "Epoch 96/100\n",
            "176/176 [==============================] - 19s 107ms/step - loss: 1.1365 - accuracy: 0.6616 - top-5-accuracy: 0.9215 - val_loss: 1.7674 - val_accuracy: 0.5440 - val_top-5-accuracy: 0.8212\n",
            "Epoch 97/100\n",
            "176/176 [==============================] - 19s 107ms/step - loss: 1.1509 - accuracy: 0.6592 - top-5-accuracy: 0.9201 - val_loss: 1.7610 - val_accuracy: 0.5476 - val_top-5-accuracy: 0.8226\n",
            "Epoch 98/100\n",
            "176/176 [==============================] - 20s 111ms/step - loss: 1.1410 - accuracy: 0.6616 - top-5-accuracy: 0.9206 - val_loss: 1.7476 - val_accuracy: 0.5440 - val_top-5-accuracy: 0.8250\n",
            "Epoch 99/100\n",
            "176/176 [==============================] - 19s 111ms/step - loss: 1.1376 - accuracy: 0.6640 - top-5-accuracy: 0.9215 - val_loss: 1.7468 - val_accuracy: 0.5524 - val_top-5-accuracy: 0.8264\n",
            "Epoch 100/100\n",
            "176/176 [==============================] - 19s 110ms/step - loss: 1.1346 - accuracy: 0.6627 - top-5-accuracy: 0.9214 - val_loss: 1.7311 - val_accuracy: 0.5572 - val_top-5-accuracy: 0.8274\n",
            "313/313 [==============================] - 5s 15ms/step - loss: 1.6747 - accuracy: 0.5604 - top-5-accuracy: 0.8311\n",
            "Test accuracy: 56.04%\n",
            "Test top 5 accuracy: 83.11%\n"
          ]
        }
      ],
      "source": [
        "def run_experiment(model):\n",
        "    optimizer = tfa.optimizers.AdamW(\n",
        "        learning_rate=learning_rate, weight_decay=weight_decay\n",
        "    )\n",
        "\n",
        "    model.compile(\n",
        "        optimizer=optimizer,\n",
        "        loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
        "        metrics=[\n",
        "            keras.metrics.SparseCategoricalAccuracy(name=\"accuracy\"),\n",
        "            keras.metrics.SparseTopKCategoricalAccuracy(5, name=\"top-5-accuracy\"),\n",
        "        ],\n",
        "    )\n",
        "\n",
        "    checkpoint_filepath = \"/tmp/checkpoint\"\n",
        "    checkpoint_callback = keras.callbacks.ModelCheckpoint(\n",
        "        checkpoint_filepath,\n",
        "        monitor=\"val_accuracy\",\n",
        "        save_best_only=True,\n",
        "        save_weights_only=True,\n",
        "    )\n",
        "\n",
        "    history = model.fit(\n",
        "        x=x_train,\n",
        "        y=y_train,\n",
        "        batch_size=batch_size,\n",
        "        epochs=num_epochs,\n",
        "        validation_split=0.1,\n",
        "        callbacks=[checkpoint_callback],\n",
        "    )\n",
        "\n",
        "    model.load_weights(checkpoint_filepath)\n",
        "    _, accuracy, top_5_accuracy = model.evaluate(x_test, y_test)\n",
        "    print(f\"Test accuracy: {round(accuracy * 100, 2)}%\")\n",
        "    print(f\"Test top 5 accuracy: {round(top_5_accuracy * 100, 2)}%\")\n",
        "\n",
        "    return history\n",
        "\n",
        "vit_classifier = create_vit_classifier()\n",
        "history = run_experiment(vit_classifier)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vMv3tHA_yaEm"
      },
      "source": [
        "- After 100 epochs, the ViT model achieves around 56.04% accuracy and 83.11% top-5 accuracy on the test data.\n",
        "- Increasing more epochs doesn't have much impact on accuracy and can implement early stopping if needed."
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": []
    },
    "environment": {
      "name": "tf2-gpu.2-4.m61",
      "type": "gcloud",
      "uri": "gcr.io/deeplearning-platform-release/tf2-gpu.2-4:m61"
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.7"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}