Deepfake_p4.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "~ This is a kaggle Notebook"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KIxA2VeH2ADr"
      },
      "source": [
        "# **DeepFake Image Detection**\n",
        "Implementing Research Paper \"Towards Universal Fake Image Detectors that generalize Across Generative Models\""
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g16st7iU19BK"
      },
      "source": [
        "# **Unzipping the data**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "76PxfqfO2fAt"
      },
      "source": [
        "Diffusion Models"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "9tmPiKKfDxA7"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "import zipfile\n",
        "zip_file_path = '/content/drive/MyDrive/Copy of diffusion_datasets.zip'\n",
        "extracted_dir = '/content/diffusion'\n",
        "with zipfile.ZipFile(zip_file_path, 'r') as zip_ref:\n",
        "    zip_ref.extractall(extracted_dir)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E2rtfIF42iSH"
      },
      "source": [
        "GAN Models"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "tnZEPBpGEZ5w"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "import zipfile\n",
        "zip_file_path = '/content/drive/MyDrive/Copy of CNN_synth_testset.zip'\n",
        "extracted_dir = '/content/gan'\n",
        "with zipfile.ZipFile(zip_file_path, 'r') as zip_ref:\n",
        "    zip_ref.extractall(extracted_dir)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3facobkS2pAw"
      },
      "source": [
        "Checking availability of GPU"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2LHEDClWDgyD",
        "outputId": "e06a929a-a4de-4615-e377-f4360a2d35ca"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "CUDA is available. Using GPU.\n"
          ]
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "import torch\n",
        "if torch.cuda.is_available():\n",
        "    device = torch.device(\"cuda\")\n",
        "    print(\"CUDA is available. Using GPU.\")\n",
        "else:\n",
        "    device = torch.device(\"cpu\")\n",
        "    print(\"CUDA is not available. Using CPU.\")\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "**Importing the libraries**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {},
      "outputs": [],
      "source": [
        "#Importing the libraries\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "import os\n",
        "import numpy as np\n",
        "from PIL import Image\n",
        "import cv2\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "07ot87jR27nQ"
      },
      "source": [
        "# **Data Reading**\n",
        "- Reading data from **ProGan** (Real and Fake Images) for **Training Phase**\n",
        "- It iterate over each file and **Reads** the image using cv2.imread()\n",
        "- It checks if the image is in **RGB Format** and appends the image if it satisfies this criterion\n",
        "- Now we **Resize** the image into **(224,224,3)** dimensions"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FHbfkYybO7A5"
      },
      "source": [
        "Use the below code to load images if you want to load the data from the unzipped file"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 60,
      "metadata": {
        "id": "S6j1S0_hE3Xo"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Inititlaizing empty lists for storing images\n",
        "real_img=[]\n",
        "fake_img=[]\n",
        "#path of folder to my training data\n",
        "folders=os.listdir(\"/content/gan/progan\")\n",
        "os.chdir(\"/content/gan/progan\")\n",
        "\n",
        "#Iterating into each folder\n",
        "for i in folders:\n",
        "    content=os.listdir(i)  #contains ['0_real,'1_fake']\n",
        "\n",
        "    path=os.path.join(i,content[0])\n",
        "    subfld=os.listdir(path)  #contains list of all images\n",
        "    for j in subfld:\n",
        "        img_path=os.path.join(i,content[0],j)\n",
        "        #Reading the image\n",
        "        img=np.array(Image.open(img_path))\n",
        "        #Checking if it is a RGB image\n",
        "        if len(np.shape(img))==3:\n",
        "            real_img.append(img)\n",
        "\n",
        "    path=os.path.join(i,content[1])\n",
        "    subfld=os.listdir(path)\n",
        "    for j in subfld:\n",
        "        img_path=os.path.join(i,content[1],j)\n",
        "        #Reading the image\n",
        "        img=np.array(Image.open(img_path))\n",
        "        #Checking if it is a RGB Image\n",
        "        if len(np.shape(img))==3:\n",
        "            fake_img.append(img)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Use the below code to load the dataset if you are loading the data from Goohle Drive\n",
        "\n",
        "Change GDrive link with your link"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "2TTpaG7p8cu1"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Initializing empty lists for storing testing images\n",
        "real_img=[]\n",
        "fake_img=[]\n",
        "#path of folder to my testing data\n",
        "os.chdir(\"/content/drive/MyDrive/progan\")\n",
        "path=\"/content/drive/MyDrive/progan/1_fake\"\n",
        "l=os.listdir(path)\n",
        "l=l[:1000]\n",
        "#Iterating into each file\n",
        "for j in l:\n",
        "    #Reading the image\n",
        "    img = cv2.imread(os.path.join(path,j))\n",
        "    #Checking the image if its RGB\n",
        "    if img is not None and len(img.shape) == 3:\n",
        "        img_resized = cv2.resize(img, (224,224))\n",
        "        fake_img.append(img_resized)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "path=\"/content/drive/MyDrive/progan/0_real\"\n",
        "l=os.listdir(path)\n",
        "l=l[:1000]\n",
        "for j in l:\n",
        "    #Reading the image\n",
        "    img = cv2.imread(os.path.join(path,j))\n",
        "    #Checking if the image is in RGB format\n",
        "    if img is not None and len(img.shape) == 3:\n",
        "        img_resized = cv2.resize(img, (224,224))\n",
        "        real_img.append(img_resized)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EruznpqrGTSn",
        "outputId": "866e7fd7-29a2-4e39-ecd8-9003247a6e6f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Real Images: (1000, 224, 224, 3)\n",
            "Fake Images (1000, 224, 224, 3)\n"
          ]
        }
      ],
      "source": [
        "print(\"Real Images:\",np.shape(real_img))\n",
        "print(\"Fake Images\",np.shape(fake_img))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JFgNyiCu557R"
      },
      "source": [
        "# **Pre-Processing of data**\n",
        "- The following function preprocesses the data using :\n",
        "    - 1) **JPEG Compression** : It converts the format of image to .jpg with 80% \n",
        "    - 2) **Gaussian Blur** :  It applies a gaussian blur filter with kernel size=(5,5) and std dev=0 to half of the images using np.random.rand()<0.5"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 61,
      "metadata": {
        "id": "jJRT-t2N5ynY"
      },
      "outputs": [],
      "source": [
        "def preprocess(orignal):\n",
        "  #JPEG comppression\n",
        "  encode_param = [int(cv2.IMWRITE_JPEG_QUALITY), 80]\n",
        "  _, comp = cv2.imencode('.jpg',orignal, encode_param)\n",
        "  comp = cv2.imdecode(comp, 1)\n",
        "\n",
        "  #Gaussian Blur\n",
        "  if np.random.rand()<0.5:\n",
        "    kernel_size = (5, 5)\n",
        "    blurred_image = cv2.GaussianBlur(orignal, kernel_size, 0)\n",
        "    return blurred_image\n",
        "  else:\n",
        "    return comp"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Showing the effect of preprocessing to any random image"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 62,
      "metadata": {
        "id": "H-_hrGQiGh21"
      },
      "outputs": [],
      "source": [
        "#Showing the effect of preprocessing to any random image\n",
        "image_to_show=cv2.imread(\"/content/drive/MyDrive/biggan/0_real/0--n01440764_5043.png\")\n",
        "image_preprocessed=preprocess(image_to_show)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 63,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 308
        },
        "id": "MyucdTliGxdU",
        "outputId": "1cff42f5-0828-40e9-e3a3-eab90c2f252a"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "#Plotting the orignal and preprocessed image\n",
        "import cv2\n",
        "import matplotlib.pyplot as plt\n",
        "plt.subplot(1, 2, 1)\n",
        "plt.imshow(cv2.cvtColor(image_to_show,cv2.COLOR_BGR2RGB))\n",
        "plt.title(\"Before\")\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "plt.imshow(cv2.cvtColor(image_preprocessed,cv2.COLOR_BGR2RGB))\n",
        "plt.title(\"After\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1EKSKtxI5VAi"
      },
      "source": [
        "Applying preprocessing to real images"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vu8Hed94G1eE",
        "outputId": "0b069d8a-c2ff-4da9-d761-27d38db30362"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(1000, 224, 224, 3)"
            ]
          },
          "execution_count": 12,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "real=[preprocess(i) for i in real_img]\n",
        "np.shape(real)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pwJ5W_xq5Z_E"
      },
      "source": [
        "Applying preprocessing to fake images"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CUcueOoVHSlQ",
        "outputId": "5a01a285-dce1-4c82-ff15-2ac3651917d5"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(1000, 224, 224, 3)"
            ]
          },
          "execution_count": 13,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "fake=[preprocess(i) for i in fake_img]\n",
        "np.shape(fake)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZMcJkgJL5gGc"
      },
      "source": [
        "# **Testing Data**\n",
        "- Reading data for **Testing Phase**\n",
        "- It iterate over each file and **Reads** the image using cv2.imread()\n",
        "- It checks if the image is in **RGB Format** and appends the image if it satisfies this criterion\n",
        "- Now we **Resize** the image into **(224,224,3)** dimensions"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "XQU6Ynz-HWyR"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Initializing empty lists for storing testing images\n",
        "r_t=[]\n",
        "f_t=[]\n",
        "#path of folder to my testing data\n",
        "os.chdir(\"/content/diffusion/diffusion_datasets\")\n",
        "path=\"/content/diffusion/diffusion_datasets/dalle/1_fake\"\n",
        "l=os.listdir(path)\n",
        "#Iterating into each file\n",
        "for j in l:\n",
        "    #Reading the image\n",
        "    img = cv2.imread(os.path.join(path,j))\n",
        "    #Checking the image if its RGB\n",
        "    if img is not None and len(img.shape) == 3:\n",
        "        img_resized = cv2.resize(img, (224,224))\n",
        "        f_t.append(img_resized)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "path=\"/content/diffusion/diffusion_datasets/laion/0_real\"\n",
        "l=os.listdir(path)\n",
        "for j in l:\n",
        "    #Reading the image\n",
        "    img = cv2.imread(os.path.join(path,j))\n",
        "    #Checking if the image is in RGB format\n",
        "    if img is not None and len(img.shape) == 3:\n",
        "        img_resized = cv2.resize(img, (224,224))\n",
        "        r_t.append(img_resized)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6HtDztZ6KFUn",
        "outputId": "c097bf7e-d64b-42a2-ee51-e6e16069ab7f"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(500, 224, 224, 3)"
            ]
          },
          "execution_count": 15,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#Slicing and taking first 500 images from each real and fake set of images\n",
        "r_tt=r_t[:500]\n",
        "f_tt=f_t[:500]\n",
        "np.shape(f_tt)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8kSayAbKLP_U",
        "outputId": "af5a4cae-d334-4dc5-abde-17732a113c91"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Testing images:\n",
            " (1000, 224, 224, 3)\n",
            "(1000,)\n"
          ]
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "test_images=np.concatenate([r_tt,f_tt])\n",
        "\n",
        "test_labels=[0]*500\n",
        "test_labels.extend([1]*500)\n",
        "\n",
        "print(\"Testing images:\\n\",np.shape(test_images))\n",
        "print(np.shape(test_labels))\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lt8d15037lhO"
      },
      "source": [
        "Applying preprocessing to the testing images"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lxuGWkH9ORY2",
        "outputId": "03a28963-db97-4fac-8445-0557adcb9483"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "(1000, 224, 224, 3)\n",
            "(1000,)\n"
          ]
        }
      ],
      "source": [
        "test=[preprocess(i) for i in test_images]\n",
        "print(np.shape(test))\n",
        "print(np.shape(test_labels))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wjJY4Gxl7un4"
      },
      "source": [
        "# **Importing CLIP Model from OpenAI Github Repository**\n",
        "\n",
        "Repo Link:https://github.com/openai/CLIP.git"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P0BD8u-GObTd",
        "outputId": "7e04e8c9-5bab-4223-aa80-3188924202ac"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Collecting ftfy\n",
            "  Downloading ftfy-6.1.3-py3-none-any.whl (53 kB)\n",
            "\u001b[?25l     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/53.4 kB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m53.4/53.4 kB\u001b[0m \u001b[31m2.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: regex in /usr/local/lib/python3.10/dist-packages (2023.12.25)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (4.66.1)\n",
            "Requirement already satisfied: wcwidth<0.3.0,>=0.2.12 in /usr/local/lib/python3.10/dist-packages (from ftfy) (0.2.13)\n",
            "Installing collected packages: ftfy\n",
            "Successfully installed ftfy-6.1.3\n",
            "Collecting git+https://github.com/openai/CLIP.git\n",
            "  Cloning https://github.com/openai/CLIP.git to /tmp/pip-req-build-lm7t36xf\n",
            "  Running command git clone --filter=blob:none --quiet https://github.com/openai/CLIP.git /tmp/pip-req-build-lm7t36xf\n",
            "  Resolved https://github.com/openai/CLIP.git to commit a1d071733d7111c9c014f024669f959182114e33\n",
            "  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: ftfy in /usr/local/lib/python3.10/dist-packages (from clip==1.0) (6.1.3)\n",
            "Requirement already satisfied: regex in /usr/local/lib/python3.10/dist-packages (from clip==1.0) (2023.12.25)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from clip==1.0) (4.66.1)\n",
            "Requirement already satisfied: torch in /usr/local/lib/python3.10/dist-packages (from clip==1.0) (2.1.0+cu121)\n",
            "Requirement already satisfied: torchvision in /usr/local/lib/python3.10/dist-packages (from clip==1.0) (0.16.0+cu121)\n",
            "Requirement already satisfied: wcwidth<0.3.0,>=0.2.12 in /usr/local/lib/python3.10/dist-packages (from ftfy->clip==1.0) (0.2.13)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from torch->clip==1.0) (3.13.1)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.10/dist-packages (from torch->clip==1.0) (4.9.0)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch->clip==1.0) (1.12)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch->clip==1.0) (3.2.1)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch->clip==1.0) (3.1.3)\n",
            "Requirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from torch->clip==1.0) (2023.6.0)\n",
            "Requirement already satisfied: triton==2.1.0 in /usr/local/lib/python3.10/dist-packages (from torch->clip==1.0) (2.1.0)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from torchvision->clip==1.0) (1.23.5)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from torchvision->clip==1.0) (2.31.0)\n",
            "Requirement already satisfied: pillow!=8.3.*,>=5.3.0 in /usr/local/lib/python3.10/dist-packages (from torchvision->clip==1.0) (9.4.0)\n",
            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch->clip==1.0) (2.1.5)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->torchvision->clip==1.0) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->torchvision->clip==1.0) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->torchvision->clip==1.0) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->torchvision->clip==1.0) (2024.2.2)\n",
            "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch->clip==1.0) (1.3.0)\n",
            "Building wheels for collected packages: clip\n",
            "  Building wheel for clip (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for clip: filename=clip-1.0-py3-none-any.whl size=1369497 sha256=c18fccbffbdefaf1a9e21563aa2cb751363256f0a517cb4b3ed2008dc6a060d3\n",
            "  Stored in directory: /tmp/pip-ephem-wheel-cache-e197i8mp/wheels/da/2b/4c/d6691fa9597aac8bb85d2ac13b112deb897d5b50f5ad9a37e4\n",
            "Successfully built clip\n",
            "Installing collected packages: clip\n",
            "Successfully installed clip-1.0\n"
          ]
        }
      ],
      "source": [
        "! pip install ftfy regex tqdm\n",
        "! pip install git+https://github.com/openai/CLIP.git"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7CemOgVG8GmJ"
      },
      "source": [
        "Taking CLIP model from Transformers\n",
        "(tried using this method to import clip model , but it was not working)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "in2hd4_J8ICF"
      },
      "outputs": [],
      "source": [
        "# from transformers import CLIPProcessor, CLIPModel\n",
        "# import torch\n",
        "\n",
        "# processor = CLIPProcessor.from_pretrained(\"openai/clip-vit-base-patch16\")\n",
        "# model = CLIPModel.from_pretrained(\"openai/clip-vit-base-patch16\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZZO_O8Ce7-dh"
      },
      "source": [
        "# **Importing the libraries**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "xwaLwBYkOfS-"
      },
      "outputs": [],
      "source": [
        "import torch\n",
        "from pkg_resources import packaging\n",
        "import torch.nn as nn\n",
        "import torchvision.transforms as transforms\n",
        "from torchvision import datasets, transforms\n",
        "from torch.utils.data import DataLoader\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.neighbors import NearestNeighbors\n",
        "from sklearn.model_selection import train_test_split"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-2DVu7TV8NW0"
      },
      "source": [
        "**Finding \"ViT-L/14\" model from CLIP**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "m5Q5kkjzOjo2",
        "outputId": "9af2f012-801d-4ae0-96fe-d35074468140"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Searching...\n",
            "Searching...\n",
            "Searching...\n",
            "Searching...\n",
            "Searching...\n",
            "Searching...\n",
            "Searching...\n",
            "Model Found\n"
          ]
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "import clip\n",
        "for i in clip.available_models():\n",
        "  if i==\"ViT-L/14\":\n",
        "      print(\"Model Found\")\n",
        "      break\n",
        "  else:\n",
        "    print(\"Searching...\")\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2NpFRKAC8ZF8"
      },
      "source": [
        "Loading the model imported above with all its parameters"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LLlU0dKwOl5X",
        "outputId": "111a32e4-aa00-44cf-faf6-d0f15420688f"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "100%|███████████████████████████████████████| 890M/890M [00:10<00:00, 89.4MiB/s]\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "CLIP(\n",
              "  (visual): VisionTransformer(\n",
              "    (conv1): Conv2d(3, 1024, kernel_size=(14, 14), stride=(14, 14), bias=False)\n",
              "    (ln_pre): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "    (transformer): Transformer(\n",
              "      (resblocks): Sequential(\n",
              "        (0): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (1): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (2): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (3): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (4): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (5): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (6): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (7): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (8): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (9): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (10): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (11): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (12): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (13): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (14): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (15): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (16): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (17): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (18): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (19): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (20): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (21): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (22): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "        (23): ResidualAttentionBlock(\n",
              "          (attn): MultiheadAttention(\n",
              "            (out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "          (mlp): Sequential(\n",
              "            (c_fc): Linear(in_features=1024, out_features=4096, bias=True)\n",
              "            (gelu): QuickGELU()\n",
              "            (c_proj): Linear(in_features=4096, out_features=1024, bias=True)\n",
              "          )\n",
              "          (ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "        )\n",
              "      )\n",
              "    )\n",
              "    (ln_post): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)\n",
              "  )\n",
              "  (transformer): Transformer(\n",
              "    (resblocks): Sequential(\n",
              "      (0): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (1): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (2): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (3): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (4): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (5): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (6): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (7): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (8): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (9): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (10): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "      (11): ResidualAttentionBlock(\n",
              "        (attn): MultiheadAttention(\n",
              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "        (mlp): Sequential(\n",
              "          (c_fc): Linear(in_features=768, out_features=3072, bias=True)\n",
              "          (gelu): QuickGELU()\n",
              "          (c_proj): Linear(in_features=3072, out_features=768, bias=True)\n",
              "        )\n",
              "        (ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              "      )\n",
              "    )\n",
              "  )\n",
              "  (token_embedding): Embedding(49408, 768)\n",
              "  (ln_final): LayerNorm((768,), eps=1e-05, elementwise_affine=True)\n",
              ")"
            ]
          },
          "execution_count": 21,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "model, preprocess_model = clip.load(\"ViT-L/14\")\n",
        "model.cuda().eval()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bhxYg--08gs-"
      },
      "source": [
        "**Preprocessing** the data from the **preprocess function of Clip model**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "id": "0E1_qYN3Op1M"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "real_t=[]\n",
        "#Iterating into each image\n",
        "for i in range(len(real)):\n",
        "  #Loading image\n",
        "  temp=Image.fromarray(real[i])\n",
        "  #Applying preprocessing function\n",
        "  real_t.append(preprocess_model(temp))\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "KtWW_O3POsmi"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "fake_t=[]\n",
        "#Iterating into each image\n",
        "for i in range(len(fake)):\n",
        "  #Loading image\n",
        "  temp=Image.fromarray(fake[i])\n",
        "  #Applying preprocessing function\n",
        "  fake_t.append(preprocess_model(temp))\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "M81Zg8U9OvBr"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "test_t=[]\n",
        "#Iterating into each image\n",
        "for i in range(len(test)):\n",
        "  #Loading image\n",
        "  temp=Image.fromarray(test[i])\n",
        "  #Applying preprocessing function\n",
        "  test_t.append(preprocess_model(temp))\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "**Stacking** all the images together and **converting** into **tensor format** using cuda"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "id": "WPPrKz6rOxVP"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "real_tr=torch.tensor(np.stack(real_t)).cuda()\n",
        "fake_tr=torch.tensor(np.stack(fake_t)).cuda()\n",
        "test_tr=torch.tensor(np.stack(test_t)).cuda()\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# **Encoding of Images into Feature Vectors**\n",
        "- Firstly it checks if GPU is available using torch.cuda.is_available(). If available, the device is set to CUDA (GPU); otherwise, it falls back to CPU.\n",
        "- The feature extraction process is performed on the image tensor using the CLIP model (model.encode_image()) with the with torch.no_grad() context manager to disable gradient tracking, as feature extraction does not require gradients."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Encoding for real Images"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "deykoD-cO2Ny",
        "outputId": "df6a0674-5887-4faa-83b8-837a137a9d32"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "torch.Size([1000, 768])"
            ]
          },
          "execution_count": 26,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Using GPU\n",
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\") \n",
        "model = model.to(device)\n",
        "real_tr = real_tr.to(device)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Encoding the image into its feature vector(1*786) using Clip Model \n",
        "with torch.no_grad():\n",
        "    features_real = model.encode_image(real_tr).float()\n",
        "features_real.shape\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Encoding for fake images"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gMppVF2aO3od",
        "outputId": "7aba1731-4b7d-4475-9f73-95dd7b26b862"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "torch.Size([1000, 768])"
            ]
          },
          "execution_count": 27,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#using GPU\n",
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
        "model = model.to(device)\n",
        "fake_tr = fake_tr.to(device)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Encoding the image into its feature vector(1*786) using Clip Model \n",
        "with torch.no_grad():\n",
        "    features_fake = model.encode_image(fake_tr).float()\n",
        "features_fake.shape\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Encoding for testing images"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZOebwVPuO5w7",
        "outputId": "6ebf0c8e-eca4-4382-cc46-91e3ab2410cf"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "torch.Size([1000, 768])"
            ]
          },
          "execution_count": 28,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Using GPU\n",
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
        "model = model.to(device)\n",
        "test_tr = test_tr.to(device)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Encoding the image into its feature vector(1*786) using Clip Model \n",
        "with torch.no_grad():\n",
        "    features_test = model.encode_image(test_tr).float()\n",
        "features_test.shape\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "**Saving** the **Encoded Images** in tensor format in a **.pt file**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 71,
      "metadata": {
        "id": "CZ2J1lkIO8QG"
      },
      "outputs": [],
      "source": [
        "torch.save(features_real, \"/content/drive/MyDrive/real.pt\")\n",
        "torch.save(features_fake, \"/content/drive/MyDrive/fake.pt\")\n",
        "torch.save(features_test, \"/content/drive/MyDrive/test_dm.pt\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# **Loading the encoded images**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "-xPhTNKlcbys"
      },
      "outputs": [],
      "source": [
        "features_fake=torch.load(\"/content/drive/MyDrive/Data/real.pt\") #Real\n",
        "features_real=torch.load(\"/content/drive/MyDrive/Data/fake.pt\") #Fake"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "**Converting** them to **numpy format** from tensors"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "y4lDLdQaPTHH"
      },
      "outputs": [],
      "source": [
        "features_real=features_real.cpu().numpy()\n",
        "features_fake=features_fake.cpu().numpy()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# **Plots**\n",
        "**Plotting a T-SNE plot for both real and fake images of ProGan Dataset**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 83,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "id": "yvRXVlj87ipP",
        "outputId": "ea26a5b5-af8a-4344-f964-2bbd93eadcdd"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Importing Libraries\n",
        "from sklearn.manifold import TSNE\n",
        "import matplotlib.pyplot as plt\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Reducing the dimensionality\n",
        "tsne=TSNE(n_components=2, random_state=40)\n",
        "t1_p=tsne.fit_transform(features_real)\n",
        "t2_p=tsne.fit_transform(features_fake)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Plotting them\n",
        "plt.figure(figsize=(8, 6))\n",
        "plt.scatter(t1_p[:, 0], t1_p[:, 1], color='blue', label='Real')\n",
        "plt.scatter(t2_p[:, 0], t2_p[:, 1], color='red', label='Fake')\n",
        "plt.legend()\n",
        "plt.title('T-SNE Visualization')\n",
        "plt.xlabel('Real')\n",
        "plt.ylabel('Fake')\n",
        "plt.show()\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# **Implementing Nearest Neighbor Approach**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "**Funcion**:\n",
        "- It finds the nearest neighbor to a given test feature vector among a set of feature vectors using cosine similarity.\n",
        "\n",
        "**Parameters**:\n",
        "- test: The test feature vector for which the nearest neighbor is to be found.\n",
        "- features: A 2D array containing feature vectors among which to find the nearest neighbor.\n",
        "\n",
        "**Returns**:\n",
        "- nn_sim: The cosine similarity between the test feature vector and its nearest neighbor."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "QrnqSvqiT8Ko"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "def find_nn(test,features):\n",
        "    cos_sim=cosine_similarity(features, [test])\n",
        "    nn_ind=np.argmax(cos_sim)\n",
        "    n_nbg = features[nn_ind]\n",
        "    nn_sim = cos_sim[nn_ind][0]\n",
        "    return nn_sim\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# **Testing Phase**\n",
        "- After training on ProGan now I am testing on **Glide-100** and **Laion**\n",
        "- The **test.pt** file is loaded and converted to a NumPy array (which contains Images from **Glide** and **Laion**)\n",
        "- A loop iterates over each feature vector in features_test.\n",
        "    - For each feature vector, the nearest neighbors are found in both the real and fake feature sets using the find_nn() function.\n",
        "    - That class which has less cosine similarity which means it is more closer to test image is assigned to the test image "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "Va0ZS-n3T9hV"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Loading the data\n",
        "features_test = torch.load(\"/content/drive/MyDrive/Data/test.pt\")\n",
        "#Converting into Numpy array\n",
        "features_test=features_test.cpu().numpy()\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#A List to store the predicted class\n",
        "predicted=[]\n",
        "for j in features_test:\n",
        "    real_nn=find_nn(j,features_real)\n",
        "    fake_nn=find_nn(j,features_fake)\n",
        "\n",
        "    if real_nn>fake_nn: \n",
        "      predicted.append(1)\n",
        "    else:\n",
        "      predicted.append(0)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Calculating accuracy"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GCLuZFE6UQX2",
        "outputId": "fd56090d-1be3-4df7-9fc7-a8e831271e58"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Accuracy:\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "75.2"
            ]
          },
          "execution_count": 25,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "test_labels=np.concatenate([[0]*500,[1]*500])\n",
        "c=0\n",
        "for i in range(len(test_labels)):\n",
        "    if(predicted[i]==test_labels[i]):\n",
        "      c=c+1\n",
        "print(\"Accuracy:\")\n",
        "(c/len(test_labels))*100\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# **Implementing Linear Classification Approach**\n",
        "\n",
        "For **Linear Classification**, I am training a neural network using TensorFlow's Keras API **(FCNN network)**\n",
        "\n",
        "1. **Data Preparation**: `features_real` and `features_fake` are concatenated into `X_train`and corresponding labels `Y_train` are defined. Then both are converted into numpy arrays\n",
        "\n",
        "2. **Hyperparameters**: \n",
        "    - **Batch Size:** 30 (as instructed in the research paper)\n",
        "    - **Epochs:** 7 (The model acheives a training accuracy of 1 so trained on less no of epochs)\n",
        "\n",
        "3. **Model Structure**: It consists of two fully connected layers: \n",
        "    - First layer has 768 units and uses the ReLU activation function.\n",
        "    - Output layer has 1 unit with a sigmoid activation function, for binary classification problems.\n",
        "\n",
        "4. **Model Compilation**: The model is compiled using the Adam optimizer and binary crossentropy loss function. \n",
        "    - Used **Adam** as optimizer\n",
        "    - Used **Binary crossentropy loss** as Loss Function\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# **Training Phase**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {},
      "outputs": [],
      "source": [
        "#importing libraries\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 119,
      "metadata": {
        "id": "aVX_yMn4P3Zr"
      },
      "outputs": [],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Preparing data\n",
        "X_train=np.concatenate([features_real,features_fake])\n",
        "Y_train=[0]*len(features_real)\n",
        "Y_train.extend([1]*len(features_test))\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Converting into numpy format\n",
        "X_train=np.array(X_train)\n",
        "Y_train=np.array(Y_train)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 120,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tswsOikKPxZu",
        "outputId": "ae35a9bd-fd31-4c7d-9573-28153556ed44"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/7\n",
            "67/67 [==============================] - 1s 3ms/step - loss: 0.1905 - accuracy: 0.9150\n",
            "Epoch 2/7\n",
            "67/67 [==============================] - 0s 3ms/step - loss: 0.0491 - accuracy: 0.9840\n",
            "Epoch 3/7\n",
            "67/67 [==============================] - 0s 3ms/step - loss: 0.0220 - accuracy: 0.9960\n",
            "Epoch 4/7\n",
            "67/67 [==============================] - 0s 3ms/step - loss: 0.0161 - accuracy: 0.9975\n",
            "Epoch 5/7\n",
            "67/67 [==============================] - 0s 4ms/step - loss: 0.0065 - accuracy: 1.0000\n",
            "Epoch 6/7\n",
            "67/67 [==============================] - 0s 3ms/step - loss: 0.0029 - accuracy: 1.0000\n",
            "Epoch 7/7\n",
            "67/67 [==============================] - 0s 3ms/step - loss: 0.0019 - accuracy: 1.0000\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7cb9687213c0>"
            ]
          },
          "execution_count": 120,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Hyperparameters\n",
        "batch_size=30\n",
        "epochs=7\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#model structure\n",
        "model_lc = Sequential()\n",
        "model_lc.add(Dense(units=768, input_shape=(768,), activation='relu'))\n",
        "model_lc.add(Dense(units=1, activation='sigmoid'))\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Training the model\n",
        "model_lc.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
        "model_lc.fit(X_train, Y_train, epochs=epochs, batch_size=batch_size)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# **Testing Phase**\n",
        "- Firstly we load the saved tensors\n",
        "- Then convert them to nunpy format\n",
        "- Then calculate the accuracy using the model above"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MPqTMZ7OWrJv"
      },
      "source": [
        "1. **Cycle GAN**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 136,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AItLPveHQZo7",
        "outputId": "4ccd9f04-6d65-41ca-f1ea-9adf9d227ca4"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Accuracy:\n",
            "32/32 [==============================] - 0s 3ms/step - loss: 0.0824 - accuracy: 0.9730\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "0.9729999899864197"
            ]
          },
          "execution_count": 136,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Loading data\n",
        "features_test=torch.load(\"/content/drive/MyDrive/test_cg.pt\")\n",
        "#Converting into tensor format\n",
        "features_test=features_test.cpu().numpy()\n",
        "#Test labels\n",
        "test_labels=np.concatenate([[0]*500,[1]*500])\n",
        "X_test=np.array(features_test)\n",
        "Y_labels=np.array(test_labels)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Testing\n",
        "print(\"Accuracy:\")\n",
        "model_lc.evaluate(X_test,Y_labels)[1]\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uQ1LvMZOXMWN"
      },
      "source": [
        "2. **Dalle-100 and Laion**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 135,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TAXAbugKRLy0",
        "outputId": "8d3b1074-0188-4dd0-fa6b-d15369d4810b"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Accuracy:\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.4469 - accuracy: 0.8690\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "0.8690000176429749"
            ]
          },
          "execution_count": 135,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Loading data\n",
        "features_test=torch.load(\"/content/drive/MyDrive/test_dm.pt\")\n",
        "features_test=features_test.cpu().numpy()\n",
        "test_labels=np.concatenate([[0]*500,[1]*500])\n",
        "X_test=np.array(features_test)\n",
        "Y_labels=np.array(test_labels)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Testing phase\n",
        "print(\"Accuracy:\")\n",
        "model_lc.evaluate(X_test,Y_labels)[1]\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C6Uht3MsYZ89"
      },
      "source": [
        "3. **Glide-100 and Laion**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 130,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qvd_PkniXOjP",
        "outputId": "f3279282-091e-40a7-ff08-d2658cae5330"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Accuracy:\n",
            "32/32 [==============================] - 0s 2ms/step - loss: 0.2380 - accuracy: 0.9290\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "0.9290000200271606"
            ]
          },
          "execution_count": 130,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "#Loading data\n",
        "features_test=torch.load(\"/content/drive/MyDrive/Data/test1.pt\")\n",
        "features_test=features_test.cpu().numpy()\n",
        "test_labels=np.concatenate([[1]*500,[0]*500])\n",
        "X_test=np.array(features_test[:1000])\n",
        "Y_labels=np.array(test_labels)\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>\n",
        "# Testing phase\n",
        "print(\"Accuracy:\")\n",
        "model_lc.evaluate(X_test,Y_labels)[1]\n",
        "#<------------------------------------------------------------------------------------------------------------------------------------------------------------------------>"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.11.0"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}