Soft_Computing_(Offline_2)_Experiment2(1st_Adjustment).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy_of_Soft_Computing_(Offline_2).ipynb",
      "provenance": [],
      "collapsed_sections": [
        "b4garZF0TMEC",
        "W7ZUVONFUrOv"
      ]
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X5-4GFDXuChh"
      },
      "source": [
        "# **Assignment 2 (Deep Neural Networks) - 160204107**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CAmqIkdlcBDx"
      },
      "source": [
        "**Adjustment - 1**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y2Bxhm2GuIDF"
      },
      "source": [
        "**Libraries**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cOFePX-pp9fi"
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import math\n",
        "\n",
        "import cv2\n",
        "import os\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms\n",
        "\n",
        "import re\n",
        "from PIL import Image\n",
        "import pickle"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XUK0viGsDhGI"
      },
      "source": [
        "## **Step - 1**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LCWQ8eX8Q8vO"
      },
      "source": [
        "### **Drive and Directories**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YUz2ADpFD2gQ"
      },
      "source": [
        "**Mount Drive**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WZlYhojBD5aM",
        "outputId": "1d2b7457-d06d-446e-d36f-4e0a162b01a4"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xIp8dq1oDsWx"
      },
      "source": [
        "**Directories**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "7LE6KS82DwC5",
        "outputId": "763d56f6-e1d6-4a71-f70b-b1b07b3c4b6b"
      },
      "source": [
        "directory = r'/content/drive/MyDrive/Soft Computing 4.2/Deep Neural Networks Dataset/'\n",
        "img_folder = r'/content/drive/MyDrive/Soft Computing 4.2/Deep Neural Networks Dataset/training-c'\n",
        "df = pd.read_csv('/content/drive/MyDrive/Soft Computing 4.2/Deep Neural Networks Dataset/training-c.csv')\n",
        "df2_Train = pd.read_csv('/content/drive/MyDrive/Soft Computing 4.2/Deep Neural Networks Dataset/verification_train.csv')\n",
        "df2_Test = pd.read_csv('/content/drive/MyDrive/Soft Computing 4.2/Deep Neural Networks Dataset/verification_test.csv')\n",
        "df"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>filename</th>\n",
              "      <th>original filename</th>\n",
              "      <th>scanid</th>\n",
              "      <th>digit</th>\n",
              "      <th>database name original</th>\n",
              "      <th>contributing team</th>\n",
              "      <th>database name</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>c00000.png</td>\n",
              "      <td>Scan_280_digit_6_num_7.png</td>\n",
              "      <td>280</td>\n",
              "      <td>6</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>c00001.png</td>\n",
              "      <td>Scan_29_digit_1_num_1.png</td>\n",
              "      <td>29</td>\n",
              "      <td>1</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>c00002.png</td>\n",
              "      <td>Scan_131_digit_3_num_1.png</td>\n",
              "      <td>131</td>\n",
              "      <td>3</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>c00003.png</td>\n",
              "      <td>Scan_160_digit_2_num_2.png</td>\n",
              "      <td>160</td>\n",
              "      <td>2</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>c00004.png</td>\n",
              "      <td>Scan_202_digit_7_num_0.png</td>\n",
              "      <td>202</td>\n",
              "      <td>7</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24293</th>\n",
              "      <td>c24293.png</td>\n",
              "      <td>Scan_167_digit_3_num_4.png</td>\n",
              "      <td>167</td>\n",
              "      <td>3</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24294</th>\n",
              "      <td>c24294.png</td>\n",
              "      <td>Scan_174_digit_2_num_8.png</td>\n",
              "      <td>174</td>\n",
              "      <td>2</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24295</th>\n",
              "      <td>c24295.png</td>\n",
              "      <td>Scan_152_digit_7_num_4.png</td>\n",
              "      <td>152</td>\n",
              "      <td>7</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24296</th>\n",
              "      <td>c24296.png</td>\n",
              "      <td>Scan_123_digit_2_num_2.png</td>\n",
              "      <td>123</td>\n",
              "      <td>2</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24297</th>\n",
              "      <td>c24297.png</td>\n",
              "      <td>Scan_2_digit_7_num_6.png</td>\n",
              "      <td>2</td>\n",
              "      <td>7</td>\n",
              "      <td>OngkoDB</td>\n",
              "      <td>Buet_Backpropers</td>\n",
              "      <td>training-c</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>24298 rows × 7 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "         filename           original filename  ...  contributing team  database name\n",
              "0      c00000.png  Scan_280_digit_6_num_7.png  ...   Buet_Backpropers     training-c\n",
              "1      c00001.png   Scan_29_digit_1_num_1.png  ...   Buet_Backpropers     training-c\n",
              "2      c00002.png  Scan_131_digit_3_num_1.png  ...   Buet_Backpropers     training-c\n",
              "3      c00003.png  Scan_160_digit_2_num_2.png  ...   Buet_Backpropers     training-c\n",
              "4      c00004.png  Scan_202_digit_7_num_0.png  ...   Buet_Backpropers     training-c\n",
              "...           ...                         ...  ...                ...            ...\n",
              "24293  c24293.png  Scan_167_digit_3_num_4.png  ...   Buet_Backpropers     training-c\n",
              "24294  c24294.png  Scan_174_digit_2_num_8.png  ...   Buet_Backpropers     training-c\n",
              "24295  c24295.png  Scan_152_digit_7_num_4.png  ...   Buet_Backpropers     training-c\n",
              "24296  c24296.png  Scan_123_digit_2_num_2.png  ...   Buet_Backpropers     training-c\n",
              "24297  c24297.png    Scan_2_digit_7_num_6.png  ...   Buet_Backpropers     training-c\n",
              "\n",
              "[24298 rows x 7 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b1jxaCh4LUZv"
      },
      "source": [
        "### **Read Dataset-1**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "nEoAt-TqFdnx",
        "outputId": "beda6d5c-6fba-4321-b06d-4d9d4305943b"
      },
      "source": [
        "df = df[['filename','digit']]\n",
        "df"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>filename</th>\n",
              "      <th>digit</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>c00000.png</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>c00001.png</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>c00002.png</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>c00003.png</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>c00004.png</td>\n",
              "      <td>7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24293</th>\n",
              "      <td>c24293.png</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24294</th>\n",
              "      <td>c24294.png</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24295</th>\n",
              "      <td>c24295.png</td>\n",
              "      <td>7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24296</th>\n",
              "      <td>c24296.png</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24297</th>\n",
              "      <td>c24297.png</td>\n",
              "      <td>7</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>24298 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "         filename  digit\n",
              "0      c00000.png      6\n",
              "1      c00001.png      1\n",
              "2      c00002.png      3\n",
              "3      c00003.png      2\n",
              "4      c00004.png      7\n",
              "...           ...    ...\n",
              "24293  c24293.png      3\n",
              "24294  c24294.png      2\n",
              "24295  c24295.png      7\n",
              "24296  c24296.png      2\n",
              "24297  c24297.png      7\n",
              "\n",
              "[24298 rows x 2 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aoBgG9eIRMVs"
      },
      "source": [
        "### **Dataset-1 Creation and Save (One Time Run)**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xUGgAwzIb95S"
      },
      "source": [
        "**Image and Digit Data Creation**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MamFtFry1dqN"
      },
      "source": [
        "img_digit = []\n",
        "dimention = (180,180)\n",
        "\n",
        "transformer=transforms.Compose([\n",
        "    transforms.Resize(dimention),\n",
        "    transforms.ToTensor(),\n",
        "    transforms.Normalize([0.5], [0.5])                         \n",
        "])\n",
        "\n",
        "for name in os.listdir(img_folder):\n",
        "    # Image name and digit from csv file\n",
        "    index = re.split('c| |\\.',name)\n",
        "    index = int(index[1])\n",
        "    digit_label = df['digit'][index]\n",
        "    \n",
        "    # Open an image using PIL library\n",
        "    image_dir = os.path.join(img_folder, name)\n",
        "    image = Image.open(image_dir)\n",
        "\n",
        "    # Tensor transformation and normalization\n",
        "    image_Ttrans = transformer(image)\n",
        "    img_digit.append((image_Ttrans,digit_label))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Oc3dSF1YCvvi"
      },
      "source": [
        "**Save Image and Digit Data in a Binary File**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4C6O-ZQHCGlJ"
      },
      "source": [
        "with open(directory+'image_digit_data', 'wb') as file:\n",
        "  pickle.dump(img_digit, file)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oesdshjxRdxP"
      },
      "source": [
        "### **Load Dataset**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nS60gFeIDMBM"
      },
      "source": [
        "**Load Image and Digit Data from a Binary File**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_Q8PGrQDDLAV"
      },
      "source": [
        "with open(directory+'image_digit_data', 'rb') as file:\n",
        "  img_digit = pickle.load(file)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "u36_d4q2LfwM"
      },
      "source": [
        "## **Step - 2**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "36BOfAEhLjyr"
      },
      "source": [
        "### **Train & Test Creation (Dataset - 1)**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t8slH0Uvij-r",
        "outputId": "a0128999-dd86-4812-bed7-7de4325dca25"
      },
      "source": [
        "df_Train, df_Test = torch.utils.data.random_split(img_digit, (round(len(img_digit)*0.8),round(len(img_digit)*0.2)), generator=torch.Generator().manual_seed(107))\n",
        "\n",
        "print('Total Data: ', len(img_digit))\n",
        "print('Train Data: ', len(df_Train))\n",
        "print('Test Data: ', len(df_Test))\n",
        "\n",
        "df_Train[0]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Total Data:  24298\n",
            "Train Data:  19438\n",
            "Test Data:  4860\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(tensor([[[0.8902, 0.9922, 1.0000,  ..., 0.9922, 1.0000, 1.0000],\n",
              "          [0.9137, 0.9922, 1.0000,  ..., 0.9922, 0.9922, 0.9922],\n",
              "          [0.9529, 0.9922, 0.9922,  ..., 1.0000, 1.0000, 0.9922],\n",
              "          ...,\n",
              "          [1.0000, 0.9922, 1.0000,  ..., 0.9922, 1.0000, 0.9922],\n",
              "          [1.0000, 1.0000, 1.0000,  ..., 1.0000, 0.9922, 0.9922],\n",
              "          [0.9922, 0.9922, 1.0000,  ..., 0.9922, 0.9922, 0.9922]]]), 0)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FSnI6DjWasfn"
      },
      "source": [
        "### **Model Creation**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4R3aFoE1I2it"
      },
      "source": [
        "#Hidden layers\n",
        "class DeepNeuralNetworkModel(nn.Module):\n",
        "    def __init__(self, input_size, num_classes, num_hidden):\n",
        "        super().__init__()\n",
        "        ### 1st hidden layer: input_size --> 200\n",
        "        self.linear_1 = nn.Linear(input_size, num_hidden)\n",
        "        ### Non-linearity in 1st hidden layer\n",
        "        self.relu_1 = nn.ELU()\n",
        "\n",
        "        ### 2nd hidden layer: 200 --> 200\n",
        "        self.linear_2 = nn.Linear(num_hidden, num_hidden)\n",
        "        ### Non-linearity in 2nd hidden layer\n",
        "        self.relu_2 = nn.ELU()\n",
        "\n",
        "        ### 3rd hidden layer: 200 --> 200\n",
        "        self.linear_3 = nn.Linear(num_hidden, num_hidden)\n",
        "        ### Non-linearity in 3rd hidden layer\n",
        "        self.relu_3 = nn.ELU()\n",
        "\n",
        "        ### 4th hidden layer: 200 --> 200\n",
        "        self.linear_4 = nn.Linear(num_hidden, num_hidden)\n",
        "        ### Non-linearity in 4th hidden layer\n",
        "        self.relu_4 = nn.ELU()\n",
        "\n",
        "        ### 5th hidden layer: 200 --> 200\n",
        "        self.linear_5 = nn.Linear(num_hidden, num_hidden)\n",
        "        ### Non-linearity in 5th hidden layer\n",
        "        self.relu_5 = nn.ELU()\n",
        "\n",
        "        ### 6th hidden layer: 200 --> 200\n",
        "        self.linear_6 = nn.Linear(num_hidden, num_hidden)\n",
        "        ### Non-linearity in 6th hidden layer\n",
        "        self.relu_6 = nn.ELU()\n",
        "\n",
        "        ### 7th hidden layer: 200 --> 200\n",
        "        self.linear_7 = nn.Linear(num_hidden, num_hidden)\n",
        "        ### Non-linearity in 7th hidden layer\n",
        "        self.relu_7 = nn.ReLU()\n",
        "\n",
        "        ### 8th hidden layer: 200 --> 200\n",
        "        self.linear_8 = nn.Linear(num_hidden, num_hidden)\n",
        "        ### Non-linearity in 8th hidden layer\n",
        "        self.relu_8 = nn.ReLU()\n",
        "\n",
        "        ### Output layer: 200 --> output_size(10)\n",
        "        self.linear_out = nn.Linear(num_hidden, num_classes)\n",
        "\n",
        "    def forward(self, x):\n",
        "        ### 1st hidden layer\n",
        "        out  = self.linear_1(x)\n",
        "        ### Non-linearity in 1st hidden layer\n",
        "        out = self.relu_1(out)\n",
        "        \n",
        "        ### 2nd hidden layer\n",
        "        out  = self.linear_2(out)\n",
        "        ### Non-linearity in 2nd hidden layer\n",
        "        out = self.relu_2(out)\n",
        "\n",
        "        ### 3rd hidden layer\n",
        "        out  = self.linear_3(out)\n",
        "        ### Non-linearity in 3rd hidden layer\n",
        "        out = self.relu_3(out)\n",
        "\n",
        "        ### 4th hidden layer\n",
        "        out  = self.linear_4(out)\n",
        "        ### Non-linearity in 4th hidden layer\n",
        "        out = self.relu_4(out)\n",
        "\n",
        "        ### 5th hidden layer\n",
        "        out  = self.linear_5(out)\n",
        "        ### Non-linearity in 5th hidden layer\n",
        "        out = self.relu_5(out)\n",
        "\n",
        "        ### 6th hidden layer\n",
        "        out  = self.linear_6(out)\n",
        "        ### Non-linearity in 6th hidden layer\n",
        "        out = self.relu_6(out)\n",
        "\n",
        "        ### 7th hidden layer\n",
        "        out  = self.linear_7(out)\n",
        "        ### Non-linearity in 7th hidden layer\n",
        "        out = self.relu_7(out)\n",
        "\n",
        "        ### 8th hidden layer\n",
        "        out  = self.linear_8(out)\n",
        "        ### Non-linearity in 8th hidden layer\n",
        "        out = self.relu_8(out)\n",
        "        \n",
        "        # Linear layer (output)\n",
        "        probas  = self.linear_out(out)\n",
        "        return probas"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YplbD1mSR3aR"
      },
      "source": [
        "### **Dataset - 1 (Model Build, Accuracy, Prediction)**\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_ZowiiPvSSeK"
      },
      "source": [
        "**Model Build**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "caRlgwe8OwFE"
      },
      "source": [
        "batch_size = 50\n",
        "num_iters = 50000\n",
        "input_dim = 180*180 # num_features = 32400\n",
        "num_hidden = 200\n",
        "output_dim = 10\n",
        "learning_rate = 0.01\n",
        "\n",
        "model = DeepNeuralNetworkModel(input_size = input_dim,\n",
        "                               num_classes = output_dim,\n",
        "                               num_hidden = num_hidden)\n",
        "\n",
        "# Device\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
        "\n",
        "# To enable GPU\n",
        "model.to(device)\n",
        "\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = torch.optim.ASGD(model.parameters(), lr=learning_rate)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xMnIvIAESWrM"
      },
      "source": [
        "**Accuracy**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p5qMslDYm2ja",
        "outputId": "5686790a-a8ad-4767-f867-2f3474575e74"
      },
      "source": [
        "loss_data,iterations = [],[]\n",
        "accuracy_data = []\n",
        "iter = 0\n",
        "\n",
        "num_epochs = math.ceil(num_iters / (len(df_Train) / batch_size))\n",
        "\n",
        "train_loader = torch.utils.data.DataLoader(dataset=df_Train, \n",
        "                                          batch_size=batch_size, \n",
        "                                          shuffle=True)   # It's better to shuffle the whole training dataset! \n",
        "\n",
        "test_loader = torch.utils.data.DataLoader(dataset=df_Test, \n",
        "                                          batch_size=batch_size, \n",
        "                                          shuffle=False) \n",
        "\n",
        "for epoch in range(num_epochs):\n",
        "  for i, (images, labels) in enumerate(train_loader):\n",
        "\n",
        "      images = images.view(-1, 180*180).to(device)\n",
        "      labels = labels.to(device)\n",
        "\n",
        "      # Clear gradients w.r.t. parameters\n",
        "      optimizer.zero_grad()\n",
        "\n",
        "      # Forward pass to get output/logits\n",
        "      outputs = model(images) \n",
        "\n",
        "      # Calculate Loss: softmax --> cross entropy loss\n",
        "      loss = criterion(outputs, labels)\n",
        "\n",
        "      # Getting gradients w.r.t. parameters\n",
        "      loss.backward()\n",
        "\n",
        "      # Updating parameters\n",
        "      optimizer.step()\n",
        "\n",
        "      iter += 1\n",
        "      if iter > num_iters:\n",
        "        break;\n",
        "\n",
        "      if iter % 500 == 0:\n",
        "          # Calculate Accuracy         \n",
        "          correct = 0\n",
        "          total = 0\n",
        "          # Iterate through test dataset\n",
        "          for images, labels in test_loader:\n",
        "              \n",
        "              images = images.view(-1, 180*180).to(device)\n",
        "\n",
        "              # Forward pass only to get logits/output\n",
        "              outputs = model(images)\n",
        "\n",
        "              # Get predictions from the maximum value\n",
        "              _, predicted = torch.max(outputs, 1)\n",
        "\n",
        "              # Total number of labels\n",
        "              total += labels.size(0)\n",
        "              # print(predicted)\n",
        "\n",
        "\n",
        "              # Total correct predictions\n",
        "              if torch.cuda.is_available():\n",
        "                  correct += (predicted.cpu() == labels.cpu()).sum() \n",
        "              else:\n",
        "                  correct += (predicted == labels).sum()\n",
        "\n",
        "          accuracy = 100 * correct.item() / total\n",
        "\n",
        "          # save the loss and iteration\n",
        "          loss_data.append(round(loss.item(),4))\n",
        "          iterations.append(iter)\n",
        "\n",
        "          # Print Loss\n",
        "          print('Iteration: {}. Loss: {}. Accuracy: {}'.format(iter, round(loss.item(),4), round(accuracy,4)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Iteration: 500. Loss: 2.3049. Accuracy: 9.7531\n",
            "Iteration: 1000. Loss: 2.3044. Accuracy: 9.7531\n",
            "Iteration: 1500. Loss: 2.3013. Accuracy: 9.7531\n",
            "Iteration: 2000. Loss: 2.3009. Accuracy: 9.7531\n",
            "Iteration: 2500. Loss: 2.3. Accuracy: 11.9342\n",
            "Iteration: 3000. Loss: 2.3004. Accuracy: 9.7531\n",
            "Iteration: 3500. Loss: 2.2999. Accuracy: 14.6914\n",
            "Iteration: 4000. Loss: 2.3001. Accuracy: 14.465\n",
            "Iteration: 4500. Loss: 2.3015. Accuracy: 11.8519\n",
            "Iteration: 5000. Loss: 2.3. Accuracy: 13.1893\n",
            "Iteration: 5500. Loss: 2.2941. Accuracy: 17.3868\n",
            "Iteration: 6000. Loss: 2.2809. Accuracy: 14.9794\n",
            "Iteration: 6500. Loss: 2.2716. Accuracy: 13.107\n",
            "Iteration: 7000. Loss: 2.1924. Accuracy: 15.3498\n",
            "Iteration: 7500. Loss: 2.2046. Accuracy: 18.1276\n",
            "Iteration: 8000. Loss: 2.2206. Accuracy: 10.6996\n",
            "Iteration: 8500. Loss: 2.1875. Accuracy: 17.0576\n",
            "Iteration: 9000. Loss: 2.1966. Accuracy: 19.4033\n",
            "Iteration: 9500. Loss: 2.0908. Accuracy: 22.6337\n",
            "Iteration: 10000. Loss: 2.2019. Accuracy: 16.6667\n",
            "Iteration: 10500. Loss: 2.1243. Accuracy: 26.7284\n",
            "Iteration: 11000. Loss: 2.0119. Accuracy: 21.4403\n",
            "Iteration: 11500. Loss: 1.959. Accuracy: 26.5844\n",
            "Iteration: 12000. Loss: 2.2575. Accuracy: 25.3704\n",
            "Iteration: 12500. Loss: 1.9552. Accuracy: 26.1317\n",
            "Iteration: 13000. Loss: 1.9505. Accuracy: 30.823\n",
            "Iteration: 13500. Loss: 1.792. Accuracy: 31.4609\n",
            "Iteration: 14000. Loss: 1.9407. Accuracy: 31.1523\n",
            "Iteration: 14500. Loss: 2.0425. Accuracy: 26.0494\n",
            "Iteration: 15000. Loss: 1.9685. Accuracy: 30.2263\n",
            "Iteration: 15500. Loss: 1.8945. Accuracy: 33.5391\n",
            "Iteration: 16000. Loss: 2.0677. Accuracy: 33.0658\n",
            "Iteration: 16500. Loss: 1.5618. Accuracy: 33.3745\n",
            "Iteration: 17000. Loss: 1.596. Accuracy: 33.5802\n",
            "Iteration: 17500. Loss: 1.7304. Accuracy: 35.5967\n",
            "Iteration: 18000. Loss: 1.6786. Accuracy: 39.8354\n",
            "Iteration: 18500. Loss: 1.776. Accuracy: 36.5638\n",
            "Iteration: 19000. Loss: 1.7991. Accuracy: 35.3292\n",
            "Iteration: 19500. Loss: 1.6608. Accuracy: 36.6255\n",
            "Iteration: 20000. Loss: 1.7554. Accuracy: 41.1111\n",
            "Iteration: 20500. Loss: 1.5653. Accuracy: 35.2263\n",
            "Iteration: 21000. Loss: 1.3335. Accuracy: 44.9383\n",
            "Iteration: 21500. Loss: 1.6905. Accuracy: 43.8683\n",
            "Iteration: 22000. Loss: 1.3996. Accuracy: 41.9342\n",
            "Iteration: 22500. Loss: 1.3122. Accuracy: 48.0041\n",
            "Iteration: 23000. Loss: 1.5105. Accuracy: 47.2634\n",
            "Iteration: 23500. Loss: 1.6879. Accuracy: 43.7654\n",
            "Iteration: 24000. Loss: 1.4201. Accuracy: 36.8724\n",
            "Iteration: 24500. Loss: 1.1941. Accuracy: 35.5967\n",
            "Iteration: 25000. Loss: 1.4883. Accuracy: 36.1523\n",
            "Iteration: 25500. Loss: 1.297. Accuracy: 43.0864\n",
            "Iteration: 26000. Loss: 1.2174. Accuracy: 48.2305\n",
            "Iteration: 26500. Loss: 1.1727. Accuracy: 50.7407\n",
            "Iteration: 27000. Loss: 1.5623. Accuracy: 51.0082\n",
            "Iteration: 27500. Loss: 1.5683. Accuracy: 42.2634\n",
            "Iteration: 28000. Loss: 1.5334. Accuracy: 49.2181\n",
            "Iteration: 28500. Loss: 1.4086. Accuracy: 49.1975\n",
            "Iteration: 29000. Loss: 1.1456. Accuracy: 60.8436\n",
            "Iteration: 29500. Loss: 1.1642. Accuracy: 58.3745\n",
            "Iteration: 30000. Loss: 1.0247. Accuracy: 57.1811\n",
            "Iteration: 30500. Loss: 1.0428. Accuracy: 51.6667\n",
            "Iteration: 31000. Loss: 1.2225. Accuracy: 55.7613\n",
            "Iteration: 31500. Loss: 0.8883. Accuracy: 60.7819\n",
            "Iteration: 32000. Loss: 0.8644. Accuracy: 63.4156\n",
            "Iteration: 32500. Loss: 0.8072. Accuracy: 65.679\n",
            "Iteration: 33000. Loss: 1.243. Accuracy: 61.1523\n",
            "Iteration: 33500. Loss: 1.368. Accuracy: 63.107\n",
            "Iteration: 34000. Loss: 0.7782. Accuracy: 65.9259\n",
            "Iteration: 34500. Loss: 1.0872. Accuracy: 65.679\n",
            "Iteration: 35000. Loss: 0.7651. Accuracy: 63.6626\n",
            "Iteration: 35500. Loss: 0.7425. Accuracy: 66.6049\n",
            "Iteration: 36000. Loss: 0.7937. Accuracy: 63.8066\n",
            "Iteration: 36500. Loss: 0.5323. Accuracy: 67.8189\n",
            "Iteration: 37000. Loss: 1.0586. Accuracy: 69.9794\n",
            "Iteration: 37500. Loss: 0.4382. Accuracy: 67.8601\n",
            "Iteration: 38000. Loss: 0.8905. Accuracy: 70.6173\n",
            "Iteration: 38500. Loss: 1.1432. Accuracy: 62.1811\n",
            "Iteration: 39000. Loss: 0.7884. Accuracy: 68.8477\n",
            "Iteration: 39500. Loss: 0.6883. Accuracy: 72.6132\n",
            "Iteration: 40000. Loss: 1.2276. Accuracy: 61.6667\n",
            "Iteration: 40500. Loss: 0.6577. Accuracy: 70.0206\n",
            "Iteration: 41000. Loss: 0.9037. Accuracy: 61.6667\n",
            "Iteration: 41500. Loss: 0.68. Accuracy: 72.9424\n",
            "Iteration: 42000. Loss: 0.6241. Accuracy: 70.823\n",
            "Iteration: 42500. Loss: 1.0312. Accuracy: 66.9959\n",
            "Iteration: 43000. Loss: 0.3898. Accuracy: 73.6626\n",
            "Iteration: 43500. Loss: 0.4438. Accuracy: 72.3251\n",
            "Iteration: 44000. Loss: 1.0342. Accuracy: 67.1193\n",
            "Iteration: 44500. Loss: 0.6135. Accuracy: 76.5021\n",
            "Iteration: 45000. Loss: 0.868. Accuracy: 70.8436\n",
            "Iteration: 45500. Loss: 0.82. Accuracy: 71.8519\n",
            "Iteration: 46000. Loss: 0.7105. Accuracy: 76.6872\n",
            "Iteration: 46500. Loss: 0.4342. Accuracy: 74.1564\n",
            "Iteration: 47000. Loss: 0.3052. Accuracy: 79.5062\n",
            "Iteration: 47500. Loss: 0.6325. Accuracy: 76.5021\n",
            "Iteration: 48000. Loss: 0.5263. Accuracy: 71.3786\n",
            "Iteration: 48500. Loss: 0.6558. Accuracy: 78.2305\n",
            "Iteration: 49000. Loss: 0.5288. Accuracy: 76.7901\n",
            "Iteration: 49500. Loss: 0.3843. Accuracy: 75.3704\n",
            "Iteration: 50000. Loss: 0.3829. Accuracy: 76.9959\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yNwkm3CuSjny"
      },
      "source": [
        "**Prediction**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 280
        },
        "id": "uqBohu_O2b-f",
        "outputId": "3dbab83f-cbed-49bc-fb98-f61287e3332f"
      },
      "source": [
        "num_predictions = 10\n",
        "\n",
        "for images, labels in test_loader:\n",
        "    break\n",
        "\n",
        "predictions = model.forward(images[0:num_predictions].view(-1, 180*180).to(device))\n",
        "predictions = torch.argmax(predictions, dim=1)\n",
        "\n",
        "print('Actual Values for Dataset 1: ')\n",
        "plt.figure(figsize=(14,14))\n",
        "for i in range(num_predictions):\n",
        "    ax=plt.subplot(1,num_predictions,i+1)\n",
        "    ax.title.set_text(labels[i].item())\n",
        "    plt.imshow(images[i].view(180, 180), cmap=plt.cm.binary)\n",
        "\n",
        "plt.show()\n",
        "\n",
        "print('Predicted Values for Dataset 1: ')\n",
        "plt.figure(figsize=(14,14))\n",
        "for i in range(num_predictions):\n",
        "    ax=plt.subplot(1,num_predictions,i+1)\n",
        "    ax.title.set_text(predictions[i].item())\n",
        "    plt.imshow(images[i].view(180, 180), cmap=plt.cm.binary)\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Actual Values for Dataset 1: \n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1008x1008 with 10 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Predicted Values for Dataset 1: \n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1008x1008 with 10 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p0K2xEH_S6YN"
      },
      "source": [
        "### **Loss Curve of Dataset -1**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 513
        },
        "id": "hd67iRRq6Zlo",
        "outputId": "2fb7fbaa-14fb-41c4-bc09-23e881bf7190"
      },
      "source": [
        "plt.figure(figsize=(12,8))\n",
        "plt.plot(iterations,loss_data,marker='o')\n",
        "plt.title('Loss vs Iterations Curve')\n",
        "plt.xlabel('Iterations')\n",
        "plt.ylabel('Loss')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mBCUxqrPRp-S"
      },
      "source": [
        "## **Step - 3**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b4garZF0TMEC"
      },
      "source": [
        "### **Train & Test Creation (Dataset - 2)**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bU0Cf-5CBfTt"
      },
      "source": [
        "# label store\n",
        "Train2_digits = np.array(df2_Train['label'])\n",
        "Test2_digits = np.array(df2_Test['label'])\n",
        "\n",
        "# label drop\n",
        "df2_Train.drop(['label'],axis=1,inplace=True)\n",
        "df2_Test.drop(['label'],axis=1,inplace=True)\n",
        "\n",
        "# Image pixel convertion into numpy(uint8 datatype)\n",
        "Train2_img = np.array(df2_Train, dtype=np.uint8)\n",
        "Test2_img = np.array(df2_Test, dtype=np.uint8)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W7ZUVONFUrOv"
      },
      "source": [
        "### **Image Data Creation (Dataset - 2)**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C8qqVAOGBjot",
        "outputId": "466baee5-f7cc-4e9a-9b27-740b04d633c6"
      },
      "source": [
        "Train2_imgTN = []\n",
        "Test2_imgTN = []\n",
        "\n",
        "transformer=transforms.Compose([\n",
        "    transforms.ToPILImage(),\n",
        "    transforms.ToTensor(),\n",
        "    transforms.Normalize([0.5], [0.5])                         \n",
        "])\n",
        "\n",
        "matrixNum = int(math.sqrt(len(df2_Test.columns)))         #features=784, Sqrt(784)=28, 28*28 image\n",
        "\n",
        "# Tensor transformation and normalization for Train Data\n",
        "for index, data in enumerate(Train2_img,start=0):\n",
        "  image = data.reshape(matrixNum,matrixNum)\n",
        "  image_Ttrans2 = transformer(image)\n",
        "  Train2_imgTN.append((image_Ttrans2,Train2_digits[index]))\n",
        "\n",
        "# Tensor transformation and normalization for Test Data\n",
        "for index, data in enumerate(Test2_img,start=0):\n",
        "  image = data.reshape(matrixNum,matrixNum)\n",
        "  image_Ttrans2 = transformer(image)\n",
        "  Test2_imgTN.append((image_Ttrans2,Test2_digits[index]))\n",
        "\n",
        "print('Train Data: ', len(Train2_imgTN))\n",
        "print('Test Data: ', len(Test2_imgTN))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Data:  60000\n",
            "Test Data:  10000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_XHfIGSrU4bm"
      },
      "source": [
        "### **Dataset - 2 (Model Build, Accuracy, Prediction)**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_dM3jRaXU6r3"
      },
      "source": [
        "**Model Build**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zrqolLk2IKiU"
      },
      "source": [
        "batch_size = 50\n",
        "num_iters = 50000\n",
        "input_dim = matrixNum*matrixNum\n",
        "num_hidden = 200\n",
        "output_dim = 10\n",
        "learning_rate = 0.01\n",
        "\n",
        "\n",
        "model2 = DeepNeuralNetworkModel(input_size = input_dim,\n",
        "                               num_classes = output_dim,\n",
        "                               num_hidden = num_hidden)\n",
        "\n",
        "# Device\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
        "\n",
        "# To enable GPU\n",
        "model2.to(device)\n",
        "\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = torch.optim.SGD(model2.parameters(), lr=learning_rate)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M28wmzznU-vi"
      },
      "source": [
        "**Accuracy**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HVFsUZO4I1k-",
        "outputId": "0a12482f-4f1a-4834-8a61-68ace17dbe0c"
      },
      "source": [
        "loss_data2,iterations2 = [],[]\n",
        "accuracy_data2 = []\n",
        "iter = 0\n",
        "\n",
        "num_epochs = math.ceil(num_iters / (len(Train2_imgTN) / batch_size))\n",
        "\n",
        "train_loader2 = torch.utils.data.DataLoader(dataset=Train2_imgTN, \n",
        "                                           batch_size=batch_size, \n",
        "                                           shuffle=True)   # It's better to shuffle the whole training dataset! \n",
        "\n",
        "test_loader2 = torch.utils.data.DataLoader(dataset=Test2_imgTN, \n",
        "                                          batch_size=batch_size, \n",
        "                                          shuffle=False) \n",
        "\n",
        "for epoch in range(num_epochs):\n",
        "    for i, (images, labels) in enumerate(train_loader2):\n",
        "\n",
        "        images = images.view(-1, matrixNum*matrixNum).to(device)\n",
        "        labels = labels.to(device)\n",
        "\n",
        "        # Clear gradients w.r.t. parameters\n",
        "        optimizer.zero_grad()\n",
        "\n",
        "        # Forward pass to get output/logits\n",
        "        outputs = model2(images) \n",
        "\n",
        "\n",
        "        # Calculate Loss: softmax --> cross entropy loss\n",
        "        loss = criterion(outputs, labels)\n",
        "\n",
        "        # Getting gradients w.r.t. parameters\n",
        "        loss.backward()\n",
        "\n",
        "        # Updating parameters\n",
        "        optimizer.step()\n",
        "\n",
        "        iter += 1\n",
        "        if iter > 20000:\n",
        "          break;\n",
        "\n",
        "        if iter % 500 == 0:\n",
        "            # Calculate Accuracy         \n",
        "            correct = 0\n",
        "            total = 0\n",
        "            # Iterate through test dataset\n",
        "            for images, labels in test_loader2:\n",
        "                images = images.view(-1, matrixNum*matrixNum).to(device)\n",
        "\n",
        "                # Forward pass only to get logits/output\n",
        "                outputs = model2(images)\n",
        "\n",
        "                # Get predictions from the maximum value\n",
        "                _, predicted = torch.max(outputs, 1)\n",
        "\n",
        "                # Total number of labels\n",
        "                total += labels.size(0)\n",
        "                # print(predicted)\n",
        "\n",
        "\n",
        "                # Total correct predictions\n",
        "                if torch.cuda.is_available():\n",
        "                    correct += (predicted.cpu() == labels.cpu()).sum() \n",
        "                else:\n",
        "                    correct += (predicted == labels).sum()\n",
        "\n",
        "            accuracy = 100 * correct.item() / total\n",
        "\n",
        "\n",
        "            # Print Loss\n",
        "            print('Iteration: {}. Loss: {}. Accuracy: {}'.format(iter, round(loss.item(),4), accuracy))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Iteration: 500. Loss: 2.2979. Accuracy: 10.1\n",
            "Iteration: 1000. Loss: 2.2698. Accuracy: 29.7\n",
            "Iteration: 1500. Loss: 1.6624. Accuracy: 20.38\n",
            "Iteration: 2000. Loss: 1.5352. Accuracy: 31.15\n",
            "Iteration: 2500. Loss: 0.9799. Accuracy: 59.36\n",
            "Iteration: 3000. Loss: 0.7304. Accuracy: 67.14\n",
            "Iteration: 3500. Loss: 0.9489. Accuracy: 65.73\n",
            "Iteration: 4000. Loss: 0.5938. Accuracy: 73.55\n",
            "Iteration: 4500. Loss: 0.4435. Accuracy: 71.92\n",
            "Iteration: 5000. Loss: 1.0062. Accuracy: 72.16\n",
            "Iteration: 5500. Loss: 0.4389. Accuracy: 76.61\n",
            "Iteration: 6000. Loss: 0.8158. Accuracy: 76.73\n",
            "Iteration: 6500. Loss: 0.6004. Accuracy: 78.75\n",
            "Iteration: 7000. Loss: 0.4424. Accuracy: 79.44\n",
            "Iteration: 7500. Loss: 0.5165. Accuracy: 80.94\n",
            "Iteration: 8000. Loss: 0.4262. Accuracy: 82.47\n",
            "Iteration: 8500. Loss: 0.464. Accuracy: 81.94\n",
            "Iteration: 9000. Loss: 0.5658. Accuracy: 79.91\n",
            "Iteration: 9500. Loss: 0.2616. Accuracy: 82.64\n",
            "Iteration: 10000. Loss: 0.7078. Accuracy: 83.3\n",
            "Iteration: 10500. Loss: 0.3296. Accuracy: 84.4\n",
            "Iteration: 11000. Loss: 0.4468. Accuracy: 84.09\n",
            "Iteration: 11500. Loss: 0.3925. Accuracy: 83.62\n",
            "Iteration: 12000. Loss: 0.4238. Accuracy: 85.19\n",
            "Iteration: 12500. Loss: 0.2357. Accuracy: 84.08\n",
            "Iteration: 13000. Loss: 0.3844. Accuracy: 84.58\n",
            "Iteration: 13500. Loss: 0.522. Accuracy: 85.07\n",
            "Iteration: 14000. Loss: 0.3523. Accuracy: 85.74\n",
            "Iteration: 14500. Loss: 0.3573. Accuracy: 84.04\n",
            "Iteration: 15000. Loss: 0.2137. Accuracy: 85.63\n",
            "Iteration: 15500. Loss: 0.3969. Accuracy: 83.81\n",
            "Iteration: 16000. Loss: 0.3106. Accuracy: 85.64\n",
            "Iteration: 16500. Loss: 0.3469. Accuracy: 84.91\n",
            "Iteration: 17000. Loss: 0.2796. Accuracy: 85.59\n",
            "Iteration: 17500. Loss: 0.2761. Accuracy: 86.02\n",
            "Iteration: 18000. Loss: 0.5333. Accuracy: 86.14\n",
            "Iteration: 18500. Loss: 0.3257. Accuracy: 85.57\n",
            "Iteration: 19000. Loss: 0.2797. Accuracy: 86.46\n",
            "Iteration: 19500. Loss: 0.1397. Accuracy: 85.96\n",
            "Iteration: 20000. Loss: 0.4346. Accuracy: 86.1\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "04wyArcMVBsa"
      },
      "source": [
        "**Prediction**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 280
        },
        "id": "eiqZaKwcKPrJ",
        "outputId": "6bb7faba-5d39-41a1-fab0-d5a87b04ea5a"
      },
      "source": [
        "num_predictions = 10\n",
        "\n",
        "for images, labels in test_loader2:\n",
        "    break\n",
        "\n",
        "predictions = model2.forward(images[0:num_predictions].view(-1, matrixNum*matrixNum).to(device))\n",
        "predictions = torch.argmax(predictions, dim=1)\n",
        "\n",
        "print('Actual Values for Dataset 2: ')\n",
        "plt.figure(figsize=(14,14))\n",
        "for i in range(num_predictions):\n",
        "    ax=plt.subplot(1,num_predictions,i+1)\n",
        "    ax.title.set_text(labels[i].item())\n",
        "    plt.imshow(images[i].view(matrixNum, matrixNum), cmap=plt.cm.binary)\n",
        "\n",
        "plt.show()\n",
        "\n",
        "print('Predicted Values for Dataset 2: ')\n",
        "plt.figure(figsize=(14,14))\n",
        "for i in range(num_predictions):\n",
        "    ax=plt.subplot(1,num_predictions,i+1)\n",
        "    ax.title.set_text(predictions[i].item())\n",
        "    plt.imshow(images[i].view(matrixNum, matrixNum), cmap=plt.cm.binary)\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Actual Values for Dataset 2: \n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1008x1008 with 10 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Predicted Values for Dataset 2: \n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1008x1008 with 10 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}