Image_Classification_using_Qiskit_approach_2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "bmdwWVPpOMDu"
   },
   "source": [
    "**Problem Statement:** There is a huge number of digital archival images of one of the influencial persons in the history of the world, Mahatma Gandhi. We want to process those images with computer vision and machine learning tools to identify the people in those images. Although classically it is an easy task. But as Havlicek V. et al. suggested in their paper (https://arxiv.org/abs/1804.11326) that we may get computational speedup if we can use efficient and effective quantum algorithm. So, our main goal of this project is to build an efficient multiclass quantum classifier to process images.\n",
    "\n",
    "**Project plan:** At first, we have chosen a publicly available dataset to run our algorithm and have found CIFAR-10 dataset is one of the suitable ones to run our experiments. The deatils of the dataset is given in the link: https://www.cs.toronto.edu/~kriz/cifar.html. From our theoretical understanding we have found that we can only perform binary classification within the hilbert space unless it is not an ensemble or any other state-of-the-art approach. So, we can only classify 2 classes at a time. Therefore, to build a multiclass classifier, we have taken 2 classes multiple times. In our project, we have three different team who have worked on 3 different platforms. Qiskit team has 3 members. IBM qiskit simulator has two different approachs. One approach is pretty straignt forward. We encode our data to the qubits, use parameterized circuit, measurement and cost function to train our quantum circuit and then with the optimized parameter from the optimizer, we performe our test on the test dataset. Another approach is given in the link: (https://github.com/Qiskit/qiskit-machine-learning/blob/stable/0.2/docs/tutorials/05_torch_connector.ipynb). In this approach, a Qiskit's TorchConnector class is introduced. TorchConnector allows a natural integration of any NeuralNetwork from Qiskit Machine Learning into a PyTorch workflow.TorchConnector takes a Qiskit NeuralNetwork and makes it available as a PyTorch Module. The resulting module can be seamlessly incorporated into PyTorch classical architectures and trained jointly without additional considerations, enabling the development and testing of novel hybrid quantum-classical machine learning architectures.We have performed our experiments on the data with both of the approachs. \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7-qqswAJbMnH"
   },
   "source": [
    "**QISKIT Platform (Approach 2):** In this approach, we has embedded a (Quantum) NeuralNetwork into a target PyTorch workflow to classify CIFAR-10 dataset in a hybrid quantum-classical manner.\n",
    "\n",
    "**Inspiration:** https://qiskit.org/textbook/ch-machine-learning/machine-learning-qiskit-pytorch.html\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WoePwSpTjpFq"
   },
   "source": [
    "**Setting up our environment:** To perform our classification task, first we need to install qiskit, pylatexenc and qiskit_machine_learning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Rji5OzHxLW1y",
    "outputId": "0e27eb15-f2f0-430c-935b-092edaadb672"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: qiskit in /usr/local/lib/python3.7/dist-packages (0.29.0)\n",
      "Requirement already satisfied: qiskit-ignis==0.6.0 in /usr/local/lib/python3.7/dist-packages (from qiskit) (0.6.0)\n",
      "Requirement already satisfied: qiskit-ibmq-provider==0.16.0 in /usr/local/lib/python3.7/dist-packages (from qiskit) (0.16.0)\n",
      "Requirement already satisfied: qiskit-aqua==0.9.4 in /usr/local/lib/python3.7/dist-packages (from qiskit) (0.9.4)\n",
      "Requirement already satisfied: qiskit-aer==0.8.2 in /usr/local/lib/python3.7/dist-packages (from qiskit) (0.8.2)\n",
      "Requirement already satisfied: qiskit-terra==0.18.1 in /usr/local/lib/python3.7/dist-packages (from qiskit) (0.18.1)\n",
      "Requirement already satisfied: scipy>=1.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-aer==0.8.2->qiskit) (1.4.1)\n",
      "Requirement already satisfied: pybind11>=2.6 in /usr/local/lib/python3.7/dist-packages (from qiskit-aer==0.8.2->qiskit) (2.7.1)\n",
      "Requirement already satisfied: numpy>=1.16.3 in /usr/local/lib/python3.7/dist-packages (from qiskit-aer==0.8.2->qiskit) (1.19.5)\n",
      "Requirement already satisfied: retworkx>=0.8.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (0.9.0)\n",
      "Requirement already satisfied: fastdtw<=0.3.4 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (0.3.4)\n",
      "Requirement already satisfied: dlx<=1.0.4 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (1.0.4)\n",
      "Requirement already satisfied: scikit-learn>=0.20.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (0.22.2.post1)\n",
      "Requirement already satisfied: yfinance<0.1.63 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (0.1.62)\n",
      "Requirement already satisfied: quandl in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (3.6.1)\n",
      "Requirement already satisfied: h5py<3.3.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (3.1.0)\n",
      "Requirement already satisfied: sympy>=1.3 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (1.7.1)\n",
      "Requirement already satisfied: setuptools>=40.1.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (57.4.0)\n",
      "Requirement already satisfied: pandas in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (1.1.5)\n",
      "Requirement already satisfied: psutil>=5 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (5.4.8)\n",
      "Requirement already satisfied: docplex>=2.21.207 in /usr/local/lib/python3.7/dist-packages (from qiskit-aqua==0.9.4->qiskit) (2.21.207)\n",
      "Requirement already satisfied: python-dateutil>=2.8.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-ibmq-provider==0.16.0->qiskit) (2.8.2)\n",
      "Requirement already satisfied: requests>=2.19 in /usr/local/lib/python3.7/dist-packages (from qiskit-ibmq-provider==0.16.0->qiskit) (2.23.0)\n",
      "Requirement already satisfied: urllib3>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from qiskit-ibmq-provider==0.16.0->qiskit) (1.24.3)\n",
      "Requirement already satisfied: websocket-client>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from qiskit-ibmq-provider==0.16.0->qiskit) (1.2.1)\n",
      "Requirement already satisfied: requests-ntlm>=1.1.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-ibmq-provider==0.16.0->qiskit) (1.1.0)\n",
      "Requirement already satisfied: symengine>0.7 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra==0.18.1->qiskit) (0.7.2)\n",
      "Requirement already satisfied: python-constraint>=1.4 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra==0.18.1->qiskit) (1.4.0)\n",
      "Requirement already satisfied: fastjsonschema>=2.10 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra==0.18.1->qiskit) (2.15.1)\n",
      "Requirement already satisfied: tweedledum<2.0,>=1.1 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra==0.18.1->qiskit) (1.1.0)\n",
      "Requirement already satisfied: jsonschema>=2.6 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra==0.18.1->qiskit) (2.6.0)\n",
      "Requirement already satisfied: ply>=3.10 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra==0.18.1->qiskit) (3.11)\n",
      "Requirement already satisfied: dill>=0.3 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra==0.18.1->qiskit) (0.3.4)\n",
      "Requirement already satisfied: six in /usr/local/lib/python3.7/dist-packages (from docplex>=2.21.207->qiskit-aqua==0.9.4->qiskit) (1.15.0)\n",
      "Requirement already satisfied: cached-property in /usr/local/lib/python3.7/dist-packages (from h5py<3.3.0->qiskit-aqua==0.9.4->qiskit) (1.5.2)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests>=2.19->qiskit-ibmq-provider==0.16.0->qiskit) (2021.5.30)\n",
      "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests>=2.19->qiskit-ibmq-provider==0.16.0->qiskit) (2.10)\n",
      "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests>=2.19->qiskit-ibmq-provider==0.16.0->qiskit) (3.0.4)\n",
      "Requirement already satisfied: ntlm-auth>=1.0.2 in /usr/local/lib/python3.7/dist-packages (from requests-ntlm>=1.1.0->qiskit-ibmq-provider==0.16.0->qiskit) (1.5.0)\n",
      "Requirement already satisfied: cryptography>=1.3 in /usr/local/lib/python3.7/dist-packages (from requests-ntlm>=1.1.0->qiskit-ibmq-provider==0.16.0->qiskit) (3.4.7)\n",
      "Requirement already satisfied: cffi>=1.12 in /usr/local/lib/python3.7/dist-packages (from cryptography>=1.3->requests-ntlm>=1.1.0->qiskit-ibmq-provider==0.16.0->qiskit) (1.14.6)\n",
      "Requirement already satisfied: pycparser in /usr/local/lib/python3.7/dist-packages (from cffi>=1.12->cryptography>=1.3->requests-ntlm>=1.1.0->qiskit-ibmq-provider==0.16.0->qiskit) (2.20)\n",
      "Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from scikit-learn>=0.20.0->qiskit-aqua==0.9.4->qiskit) (1.0.1)\n",
      "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.7/dist-packages (from sympy>=1.3->qiskit-aqua==0.9.4->qiskit) (1.2.1)\n",
      "Requirement already satisfied: multitasking>=0.0.7 in /usr/local/lib/python3.7/dist-packages (from yfinance<0.1.63->qiskit-aqua==0.9.4->qiskit) (0.0.9)\n",
      "Requirement already satisfied: lxml>=4.5.1 in /usr/local/lib/python3.7/dist-packages (from yfinance<0.1.63->qiskit-aqua==0.9.4->qiskit) (4.6.3)\n",
      "Requirement already satisfied: pytz>=2017.2 in /usr/local/lib/python3.7/dist-packages (from pandas->qiskit-aqua==0.9.4->qiskit) (2018.9)\n",
      "Requirement already satisfied: more-itertools in /usr/local/lib/python3.7/dist-packages (from quandl->qiskit-aqua==0.9.4->qiskit) (8.8.0)\n",
      "Requirement already satisfied: inflection>=0.3.1 in /usr/local/lib/python3.7/dist-packages (from quandl->qiskit-aqua==0.9.4->qiskit) (0.5.1)\n",
      "Requirement already satisfied: pylatexenc in /usr/local/lib/python3.7/dist-packages (2.10)\n"
     ]
    }
   ],
   "source": [
    "!pip install qiskit\n",
    "!pip install pylatexenc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "qozntkFcMJuW",
    "outputId": "cfb9b274-12ef-42f3-b811-8e47120be70b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: qiskit_machine_learning in /usr/local/lib/python3.7/dist-packages (0.2.0)\n",
      "Requirement already satisfied: fastdtw in /usr/local/lib/python3.7/dist-packages (from qiskit_machine_learning) (0.3.4)\n",
      "Requirement already satisfied: qiskit-terra>=0.18.0 in /usr/local/lib/python3.7/dist-packages (from qiskit_machine_learning) (0.18.1)\n",
      "Requirement already satisfied: psutil>=5 in /usr/local/lib/python3.7/dist-packages (from qiskit_machine_learning) (5.4.8)\n",
      "Requirement already satisfied: scikit-learn>=0.20.0 in /usr/local/lib/python3.7/dist-packages (from qiskit_machine_learning) (0.22.2.post1)\n",
      "Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.7/dist-packages (from qiskit_machine_learning) (1.19.5)\n",
      "Requirement already satisfied: scipy>=1.4 in /usr/local/lib/python3.7/dist-packages (from qiskit_machine_learning) (1.4.1)\n",
      "Requirement already satisfied: setuptools>=40.1.0 in /usr/local/lib/python3.7/dist-packages (from qiskit_machine_learning) (57.4.0)\n",
      "Requirement already satisfied: python-constraint>=1.4 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (1.4.0)\n",
      "Requirement already satisfied: ply>=3.10 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (3.11)\n",
      "Requirement already satisfied: python-dateutil>=2.8.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (2.8.2)\n",
      "Requirement already satisfied: sympy>=1.3 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (1.7.1)\n",
      "Requirement already satisfied: symengine>0.7 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (0.7.2)\n",
      "Requirement already satisfied: dill>=0.3 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (0.3.4)\n",
      "Requirement already satisfied: jsonschema>=2.6 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (2.6.0)\n",
      "Requirement already satisfied: fastjsonschema>=2.10 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (2.15.1)\n",
      "Requirement already satisfied: tweedledum<2.0,>=1.1 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (1.1.0)\n",
      "Requirement already satisfied: retworkx>=0.9.0 in /usr/local/lib/python3.7/dist-packages (from qiskit-terra>=0.18.0->qiskit_machine_learning) (0.9.0)\n",
      "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.7/dist-packages (from python-dateutil>=2.8.0->qiskit-terra>=0.18.0->qiskit_machine_learning) (1.15.0)\n",
      "Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from scikit-learn>=0.20.0->qiskit_machine_learning) (1.0.1)\n",
      "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.7/dist-packages (from sympy>=1.3->qiskit-terra>=0.18.0->qiskit_machine_learning) (1.2.1)\n"
     ]
    }
   ],
   "source": [
    "!pip install qiskit_machine_learning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "uropJFQ_kI_v"
   },
   "source": [
    "**Importing Required Libraries:** We have to import all the reuired libraries and functions first to download, manipulate, train-test and optimize the quantum system to reach the desired outcome."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "YnDJmTtDLsZQ"
   },
   "outputs": [],
   "source": [
    "from torch import cat, no_grad\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision import datasets, transforms\n",
    "import torch.optim as optim\n",
    "from torch.nn import (Module, Conv2d, Linear, Dropout2d, NLLLoss,\n",
    "                     MaxPool2d, Flatten, Sequential, ReLU)\n",
    "import torch.nn.functional as F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "AndU5WAqMKiM"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from torch import Tensor\n",
    "from torch.nn import Linear, CrossEntropyLoss, MSELoss\n",
    "from torch.optim import LBFGS\n",
    "import torch\n",
    "import torchvision\n",
    "import torchvision.transforms as transforms\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from qiskit  import Aer, QuantumCircuit\n",
    "from qiskit.utils import QuantumInstance\n",
    "from qiskit.opflow import AerPauliExpectation\n",
    "from qiskit.circuit import Parameter\n",
    "from qiskit.circuit.library import RealAmplitudes, ZZFeatureMap\n",
    "from qiskit_machine_learning.neural_networks import CircuitQNN, TwoLayerQNN\n",
    "from qiskit_machine_learning.connectors import TorchConnector"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "aN9Ke-zakx75"
   },
   "source": [
    "**Data Accumulation and Defining Dataloader:** We take advantage of the torchvision API to directly load CIFAR10 dataset and define torch DataLoaders for train and test. We have downloaded our data and assigned it to X_train and X_test. We also performed *Grayscale* transformation and image resizing to adjust our image to do further processing. Then, we define our classes. We checked the data size, shape and labels. After that, as promised earlier, we took two of the ten classes from the data to do classification. In this experiement, we have chosen first two classes, Airplane and Automobile."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "mrjIZ68BMWJd",
    "outputId": "48c3956c-4c8a-4c8d-fead-cae0ab79dc51"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Files already downloaded and verified\n"
     ]
    }
   ],
   "source": [
    "batch_size = 1\n",
    "import torchvision as tv\n",
    "import torch.utils.data as data\n",
    "\n",
    "trainTransform  = tv.transforms.Compose([tv.transforms.Grayscale(num_output_channels=1),transforms.Resize(size=(28, 28)),\n",
    "                                    tv.transforms.ToTensor(), \n",
    "                                    transforms.Normalize(mean=[0.5], std=[0.5])])\n",
    "X_train        = tv.datasets.CIFAR10(root='./data', train=True, download=True, transform=trainTransform)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "TU88638zm9AM"
   },
   "source": [
    "**Assigning Classes:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "_cZadW5jM2-O"
   },
   "outputs": [],
   "source": [
    "classes=['airplane','automobile','bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "aJzYqdUunOas"
   },
   "source": [
    "**Data Pattern Recognition:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "QI-rnEXrUPcO",
    "outputId": "b11e8c27-1480-4ecf-f508-e2d33c275daf"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "50000"
      ]
     },
     "execution_count": 8,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "TIYh_LCFkJ85",
    "outputId": "93e333b7-6dc2-4aae-f474-133e99abf0f3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(tensor([[[-0.6000, -0.6941, -0.6392, -0.4980, -0.3176, -0.1686, -0.1137,\n",
       "           -0.0745, -0.0980, -0.2078, -0.1686, -0.1216, -0.1529, -0.2000,\n",
       "           -0.1608, -0.1922, -0.1922, -0.1137, -0.1451, -0.1608, -0.1373,\n",
       "           -0.0510,  0.0275,  0.0353,  0.0196,  0.0196, -0.0196, -0.0353],\n",
       "          [-0.8667, -0.9529, -0.8039, -0.5922, -0.3647, -0.2471, -0.2706,\n",
       "           -0.2863, -0.3412, -0.3882, -0.3961, -0.3569, -0.3490, -0.3333,\n",
       "           -0.3098, -0.3647, -0.3412, -0.2471, -0.2549, -0.2314, -0.2706,\n",
       "           -0.2863, -0.2314, -0.2078, -0.1843, -0.2392, -0.2863, -0.2941],\n",
       "          [-0.8196, -0.8275, -0.6078, -0.4353, -0.3098, -0.2706, -0.3412,\n",
       "           -0.3569, -0.3176, -0.3255, -0.3647, -0.2549, -0.2314, -0.2471,\n",
       "           -0.2863, -0.3020, -0.2941, -0.2549, -0.2000, -0.1686, -0.2235,\n",
       "           -0.2549, -0.2392, -0.2000, -0.1843, -0.2627, -0.2784, -0.3882],\n",
       "          [-0.7569, -0.6784, -0.4353, -0.3569, -0.3490, -0.3412, -0.3882,\n",
       "           -0.3725, -0.2627, -0.3020, -0.3725, -0.2627, -0.2549, -0.3333,\n",
       "           -0.3882, -0.3333, -0.2863, -0.2627, -0.1765, -0.1765, -0.1922,\n",
       "           -0.2392, -0.2549, -0.2706, -0.2941, -0.3647, -0.4118, -0.5216],\n",
       "          [-0.6314, -0.5451, -0.3569, -0.2784, -0.2863, -0.2941, -0.3412,\n",
       "           -0.3725, -0.2784, -0.3412, -0.4039, -0.3569, -0.3490, -0.4196,\n",
       "           -0.4510, -0.4196, -0.4275, -0.3804, -0.3176, -0.3647, -0.2706,\n",
       "           -0.1922, -0.2392, -0.3333, -0.4196, -0.5529, -0.7176, -0.7647],\n",
       "          [-0.4588, -0.3804, -0.2941, -0.2549, -0.2549, -0.3098, -0.3412,\n",
       "           -0.3020, -0.2392, -0.3255, -0.3647, -0.4667, -0.5373, -0.5686,\n",
       "           -0.4980, -0.5137, -0.6235, -0.6078, -0.5451, -0.3333, -0.3098,\n",
       "           -0.1922, -0.1608, -0.3176, -0.4510, -0.5059, -0.5608, -0.6549],\n",
       "          [-0.3255, -0.3098, -0.2235, -0.2157, -0.2706, -0.2549, -0.2549,\n",
       "           -0.2627, -0.2314, -0.2941, -0.4902, -0.6706, -0.7490, -0.6471,\n",
       "           -0.1137, -0.0667, -0.4667, -0.6471, -0.5843, -0.3647, -0.3882,\n",
       "           -0.2392, -0.1451, -0.2627, -0.3804, -0.4196, -0.3176, -0.3804],\n",
       "          [-0.1608, -0.2392, -0.2157, -0.2627, -0.2941, -0.2706, -0.2549,\n",
       "           -0.2706, -0.2078, -0.3961, -0.6706, -0.7176, -0.7176, -0.5843,\n",
       "           -0.0588, -0.0353, -0.2784, -0.5059, -0.5843, -0.5922, -0.6078,\n",
       "           -0.3882, -0.1922, -0.2078, -0.2392, -0.2784, -0.2706, -0.2235],\n",
       "          [-0.0118, -0.0353, -0.0745, -0.1843, -0.2314, -0.2314, -0.2157,\n",
       "           -0.1686, -0.2941, -0.5686, -0.6471, -0.6157, -0.4980, -0.4196,\n",
       "           -0.2863, -0.3098, -0.2706, -0.3412, -0.3490, -0.4118, -0.5529,\n",
       "           -0.4118, -0.2235, -0.2157, -0.2078, -0.2078, -0.2471, -0.2235],\n",
       "          [-0.0275, -0.0510, -0.0824, -0.1608, -0.1843, -0.1922, -0.2314,\n",
       "           -0.1451, -0.3490, -0.4980, -0.4353, -0.3647, -0.1765, -0.1922,\n",
       "           -0.1529, -0.2235, -0.1529, -0.0902,  0.1216,  0.0431, -0.3412,\n",
       "           -0.3882, -0.2157, -0.2235, -0.2000, -0.2000, -0.2392, -0.2157],\n",
       "          [-0.0667, -0.1059, -0.1294, -0.1529, -0.2000, -0.3490, -0.4196,\n",
       "           -0.2784, -0.3804, -0.2314, -0.0667,  0.0118,  0.1608,  0.0667,\n",
       "            0.1451,  0.1216,  0.1451,  0.1686,  0.2863,  0.1529, -0.2314,\n",
       "           -0.3412, -0.2078, -0.2078, -0.2000, -0.2157, -0.2314, -0.2157],\n",
       "          [ 0.0039, -0.0745, -0.1216, -0.1451, -0.3020, -0.5216, -0.5294,\n",
       "           -0.4824, -0.3882,  0.0353,  0.2157,  0.1843,  0.4902,  0.4353,\n",
       "            0.4431,  0.4196,  0.3647,  0.3333,  0.3098,  0.2157, -0.0039,\n",
       "           -0.1373, -0.2078, -0.2078, -0.2157, -0.2000, -0.2157, -0.2000],\n",
       "          [-0.0353, -0.0745, -0.1137, -0.1608, -0.4118, -0.5373, -0.4667,\n",
       "           -0.4980, -0.2000,  0.2078,  0.0980,  0.1608,  0.5373,  0.7020,\n",
       "            0.7020,  0.6471,  0.5922,  0.5059,  0.5373,  0.5373,  0.4980,\n",
       "            0.3176, -0.1137, -0.2157, -0.2078, -0.1922, -0.2235, -0.2157],\n",
       "          [-0.0667, -0.1373, -0.1451, -0.1765, -0.4588, -0.6000, -0.5137,\n",
       "           -0.3176,  0.2392,  0.2549, -0.0980,  0.1529,  0.3725,  0.7725,\n",
       "            0.9294,  0.8824,  0.7569,  0.6549,  0.5765,  0.6471,  0.8275,\n",
       "            0.6471, -0.0275, -0.2078, -0.2078, -0.2314, -0.2471, -0.2549],\n",
       "          [-0.0824, -0.1843, -0.1608, -0.1843, -0.4667, -0.5216, -0.3569,\n",
       "            0.1294,  0.5765,  0.2000,  0.0118,  0.2784,  0.4824,  0.7490,\n",
       "            0.8275,  0.7569,  0.6706,  0.5373,  0.2392,  0.3882,  0.7961,\n",
       "            0.5922,  0.0588, -0.0431, -0.0588, -0.0980, -0.2000, -0.2863],\n",
       "          [-0.1059, -0.2471, -0.1843, -0.1843, -0.3333, -0.3961, -0.0431,\n",
       "            0.5686,  0.6941,  0.1216,  0.2863,  0.5843,  0.5843,  0.5922,\n",
       "            0.5059,  0.3020,  0.2000,  0.2627,  0.3098,  0.3412,  0.6549,\n",
       "            0.5451, -0.0039, -0.1529, -0.1686, -0.1373, -0.1137, -0.1373],\n",
       "          [-0.1137, -0.3176, -0.2706, -0.2314, -0.2471, -0.2392,  0.3725,\n",
       "            0.8196,  0.6314,  0.0980,  0.3490,  0.5922,  0.2392,  0.0980,\n",
       "            0.0039, -0.0353, -0.0353,  0.0510,  0.3647,  0.6000,  0.7333,\n",
       "            0.5451, -0.0902, -0.3098, -0.3333, -0.3490, -0.2863, -0.0667],\n",
       "          [-0.1529, -0.2784, -0.2392, -0.2392, -0.2000,  0.2157,  0.7333,\n",
       "            0.7255,  0.4196,  0.0431,  0.1529,  0.3020, -0.1059, -0.1686,\n",
       "           -0.1451, -0.0039, -0.0980, -0.1216,  0.1843,  0.5451,  0.5843,\n",
       "            0.4196, -0.1137, -0.3725, -0.3490, -0.3490, -0.3647, -0.1216],\n",
       "          [-0.1608, -0.3020, -0.2627, -0.2784, -0.1608,  0.1216,  0.3804,\n",
       "            0.4039,  0.2471,  0.0353,  0.0824,  0.2235, -0.0510, -0.1608,\n",
       "           -0.0980, -0.0510, -0.1529, -0.1608, -0.0667,  0.0824,  0.0745,\n",
       "            0.0745, -0.1294, -0.2941, -0.2235, -0.1843, -0.1922,  0.0588],\n",
       "          [-0.1686, -0.2863, -0.2392, -0.2392, -0.1294, -0.1451, -0.0824,\n",
       "            0.1686,  0.1765,  0.2157,  0.2549,  0.1843, -0.0275, -0.1608,\n",
       "           -0.1373, -0.1059, -0.1059, -0.1294, -0.1373, -0.1216, -0.1373,\n",
       "           -0.1765, -0.1686, -0.1373, -0.1294, -0.1451, -0.0588,  0.2941],\n",
       "          [-0.1451, -0.2392, -0.2078, -0.1294, -0.0980, -0.1216, -0.1137,\n",
       "            0.0353,  0.0902,  0.2627,  0.3882,  0.3725,  0.0667, -0.1373,\n",
       "           -0.1529, -0.1686, -0.0824, -0.0824, -0.1373, -0.2941, -0.4196,\n",
       "           -0.4118, -0.1922, -0.0431, -0.1373, -0.2078,  0.0196,  0.2863],\n",
       "          [-0.0980, -0.2627, -0.2078, -0.1451, -0.1137, -0.0745, -0.0196,\n",
       "            0.0902,  0.1451,  0.1373,  0.4039,  0.4039,  0.2706,  0.1529,\n",
       "           -0.1294, -0.2549, -0.1451, -0.0980, -0.2549, -0.4745, -0.4902,\n",
       "           -0.3412, -0.1059, -0.0980, -0.2078, -0.1137,  0.2078,  0.0667],\n",
       "          [ 0.0745, -0.1765, -0.2392, -0.2549, -0.2314, -0.1294, -0.0431,\n",
       "           -0.0039,  0.0353, -0.0824,  0.0039,  0.0510, -0.0275, -0.0196,\n",
       "           -0.1294, -0.1922, -0.2235, -0.2471, -0.2392, -0.1843, -0.1216,\n",
       "           -0.1059, -0.1451, -0.2157, -0.1765,  0.1451,  0.3569, -0.0902],\n",
       "          [ 0.2706,  0.0196, -0.1529, -0.2549, -0.2706, -0.2000, -0.1373,\n",
       "           -0.1373, -0.1922, -0.2706, -0.1922, -0.0745, -0.1373, -0.2471,\n",
       "           -0.2392, -0.1137, -0.1294, -0.1843, -0.1529, -0.0902, -0.0980,\n",
       "           -0.0824, -0.1451, -0.2000, -0.0118,  0.3176,  0.1529, -0.3255],\n",
       "          [ 0.3725,  0.1137,  0.0275, -0.0824, -0.2000, -0.1922, -0.1216,\n",
       "           -0.1451, -0.2157, -0.2706, -0.1843, -0.1451, -0.2000, -0.2706,\n",
       "           -0.2706, -0.1765, -0.1059, -0.1294, -0.2235, -0.2157, -0.1373,\n",
       "           -0.1059, -0.1137, -0.1843,  0.0431,  0.2706, -0.2706, -0.6706],\n",
       "          [ 0.3569,  0.2000,  0.1765,  0.0980,  0.0039, -0.0667, -0.0980,\n",
       "           -0.1686, -0.2235, -0.2000, -0.1608, -0.1529, -0.2078, -0.2471,\n",
       "           -0.2314, -0.1529, -0.1137, -0.1059, -0.1922, -0.2941, -0.2392,\n",
       "           -0.1373, -0.1843, -0.4275, -0.2863,  0.0510, -0.4980, -0.7412],\n",
       "          [ 0.1765,  0.0902,  0.1686,  0.1922,  0.1922,  0.1451,  0.0275,\n",
       "           -0.1137, -0.2000, -0.1686, -0.1686, -0.1843, -0.2000, -0.2157,\n",
       "           -0.1843, -0.1529, -0.1529, -0.1608, -0.2157, -0.3098, -0.3412,\n",
       "           -0.2863, -0.4275, -0.7176, -0.5216,  0.0275, -0.3412, -0.5686],\n",
       "          [ 0.1608,  0.0902,  0.1686,  0.2392,  0.3725,  0.4118,  0.3333,\n",
       "            0.2000,  0.0745,  0.0275,  0.0118, -0.0039, -0.0588, -0.0353,\n",
       "           -0.0118, -0.0745, -0.1294, -0.1059, -0.1529, -0.2549, -0.2706,\n",
       "           -0.2314, -0.3412, -0.4667, -0.1843,  0.2941,  0.0510, -0.2471]]]),\n",
       " 6)"
      ]
     },
     "execution_count": 9,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "UNhBoh4DnmPC"
   },
   "source": [
    "**Selecting 2 Classes from 10:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "itaenR2gNFet"
   },
   "outputs": [],
   "source": [
    "idx= []\n",
    "for i in range(int(0.1*len(X_train))):\n",
    "  if X_train[i][1] in [0,1]:\n",
    "        idx.append(X_train[i])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YXENiCvznvCD"
   },
   "source": [
    "**Data Pattern Recognition for 2 Classes:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "qINH11qbNIGm",
    "outputId": "d28841ba-9f59-4e5d-94a4-e9136433f3c8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "965"
      ]
     },
     "execution_count": 11,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(idx)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-b71k_bHn5DO"
   },
   "source": [
    "**Defining Data Loader For 2 Classes:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "4lqa9xMqNV2N",
    "outputId": "2c826e9f-165b-4ef8-8957-0543c734cf80"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 1, 28, 28])\n"
     ]
    }
   ],
   "source": [
    "trainloader= data.DataLoader(idx, batch_size=batch_size, shuffle=True, num_workers=2)\n",
    "images, labels  = iter(trainloader).next()\n",
    "print (images.size())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "L1zZ_zngohYE"
   },
   "source": [
    "**Data Visualization:** We also need to visualize our data to ensure whether we are on right track. Data Visualization also helps communicate our understanding of data in simple way. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 283
    },
    "id": "CdxNKluKNncW",
    "outputId": "7d2f672a-9fb4-4155-d950-30df9d2d8fb1"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "airplane\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "def imshow(img):\n",
    "    img = img / 2 + 0.5\n",
    "    npimg = img.numpy()\n",
    "    plt.imshow(np.transpose(npimg, (1, 2, 0)))\n",
    "    plt.show()\n",
    "\n",
    "dataiter = iter(trainloader)\n",
    "images, labels = dataiter.next()\n",
    "\n",
    "imshow(torchvision.utils.make_grid(images))\n",
    "print(' '.join('%5s' % classes[labels[j]] for j in range(batch_size)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "uWSr1HRopr6V"
   },
   "source": [
    "**Plotting Several Other Training Images:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 130
    },
    "id": "WMmIeAAUNrTi",
    "outputId": "73568f09-5686-4579-c63d-ce890f32eaba"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 6 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_samples_show = 6 #six samples chosen\n",
    "\n",
    "data_iter = iter(trainloader)\n",
    "fig, axes = plt.subplots(nrows=1, ncols=n_samples_show, figsize=(10, 3))\n",
    "\n",
    "while n_samples_show > 0:\n",
    "    images, targets = data_iter.__next__()\n",
    "\n",
    "    axes[n_samples_show - 1].imshow(images[0, 0].numpy().squeeze(), cmap='gray')\n",
    "    axes[n_samples_show - 1].set_xticks([])\n",
    "    axes[n_samples_show - 1].set_yticks([])\n",
    "    axes[n_samples_show - 1].set_title(\"Labeled: {}\".format(targets[0].item()))\n",
    "\n",
    "    n_samples_show -= 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MrVkOgW1qAhD"
   },
   "source": [
    "**Defining Dataloader For Test Dataset:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "5WQBr0z-Oabl",
    "outputId": "e398a7a3-9b43-45cc-e83f-a78d608ff1b4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Files already downloaded and verified\n",
      "torch.Size([1, 1, 28, 28])\n"
     ]
    }
   ],
   "source": [
    "testTransform  = tv.transforms.Compose([tv.transforms.Grayscale(num_output_channels=1),transforms.Resize(size=(28, 28)),\n",
    "                                    tv.transforms.ToTensor(), \n",
    "                                    transforms.Normalize(mean=[0.5], std=[0.5])])\n",
    "X_test = datasets.CIFAR10(root='./data', train=False, download=True,\n",
    "                        transform=testTransform)\n",
    "\n",
    "idx_test= []\n",
    "for i in range(int(0.05*len(X_train))):\n",
    "  if X_test[i][1] in [0,1]:\n",
    "        idx_test.append(X_test[i])\n",
    "\n",
    "testloader= data.DataLoader(idx_test, batch_size=batch_size, shuffle=True, num_workers=2)\n",
    "images, labels  = iter(testloader).next()\n",
    "print (images.size())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8LUbkTgnqKZ0"
   },
   "source": [
    "**Verifying Data Size and Data Pattern:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "nHwquP_SmV2M",
    "outputId": "5465a98c-c305-4140-9034-13e8ab7cf535"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000\n"
     ]
    }
   ],
   "source": [
    "print(len(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "vXKXdndbW4Ik",
    "outputId": "d6cd071d-66a8-4e20-9b44-40836eefa7ab"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "493"
      ]
     },
     "execution_count": 19,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(idx_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "j05KuYBFqXJN"
   },
   "source": [
    "**Defining the QNN and Hybrid Model:** This is where we can see the power of the TorchConnector. First, We created our quantum neural network layer(TwoLayerQNN). Then, we embed it into a layer in our torch Module by initializing a torch connector as TorchConnector(qnn)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "Wf2NU94iMTaN"
   },
   "outputs": [],
   "source": [
    "qi = QuantumInstance(Aer.get_backend('statevector_simulator'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "9vIzlwONW7uT",
    "outputId": "266a08a2-a7cc-4b50-a3ae-964ed5591fbf"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ComposedOp([\n",
      "  OperatorMeasurement(1.0 * ZZ),\n",
      "  CircuitStateFn(\n",
      "       ┌──────────────────────────┐»\n",
      "  q_0: ┤0                         ├»\n",
      "       │  ZZFeatureMap(x[0],x[1]) │»\n",
      "  q_1: ┤1                         ├»\n",
      "       └──────────────────────────┘»\n",
      "  «     ┌────────────────────────────────────────────────┐\n",
      "  «q_0: ┤0                                               ├\n",
      "  «     │  RealAmplitudes(θ[0],θ[1],θ[2],θ[3],θ[4],θ[5]) │\n",
      "  «q_1: ┤1                                               ├\n",
      "  «     └────────────────────────────────────────────────┘\n",
      "  )\n",
      "])\n"
     ]
    }
   ],
   "source": [
    "feature_map = ZZFeatureMap(2)\n",
    "ansatz = RealAmplitudes(2, reps=2)\n",
    "qnn4 = TwoLayerQNN(2, feature_map, ansatz, input_gradients=True, exp_val=AerPauliExpectation(), quantum_instance=qi)\n",
    "print(qnn4.operator)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "id": "PddiM37YXP5c"
   },
   "outputs": [],
   "source": [
    "class Net(Module):\n",
    "\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.conv1 = Conv2d(1, 2, kernel_size=5)\n",
    "        self.conv2 = Conv2d(2, 16, kernel_size=5)\n",
    "        self.dropout = Dropout2d()\n",
    "        self.fc1 = Linear(256, 64)\n",
    "        self.fc2 = Linear(64, 2)         \n",
    "        self.qnn = TorchConnector(qnn4)  \n",
    "        self.fc3 = Linear(1, 1)          \n",
    "\n",
    "    def forward(self, x):\n",
    "        x = F.relu(self.conv1(x))\n",
    "        x = F.max_pool2d(x, 2)\n",
    "        x = F.relu(self.conv2(x))\n",
    "        x = F.max_pool2d(x, 2)\n",
    "        x = self.dropout(x)\n",
    "        x = x.view(1, -1)\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = self.fc2(x)\n",
    "        x = self.qnn(x)  # apply QNN\n",
    "        x = self.fc3(x)\n",
    "        return cat((x, 1 - x), -1)\n",
    "\n",
    "model4 = Net()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "3tZg9MR-s7aa"
   },
   "source": [
    "**Training Hybrid QNN Model:** In this section, we have trained our QNN. We have used ADAM optimizer for optimizaion. Learning rate is 0.001 and total number of epochs is 10. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "w83iL3joXjbU",
    "outputId": "142e1dd6-8b68-4c71-9eac-b18e8da12a3d"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:718: UserWarning: Named tensors and all their associated APIs are an experimental feature and subject to change. Please do not use them for anything important until they are released as stable. (Triggered internally at  /pytorch/c10/core/TensorImpl.h:1156.)\n",
      "  return torch.max_pool2d(input, kernel_size, stride, padding, dilation, ceil_mode)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training [10%]\tLoss: -0.4831\n",
      "Training [20%]\tLoss: -0.7279\n",
      "Training [30%]\tLoss: -1.0307\n",
      "Training [40%]\tLoss: -1.4283\n",
      "Training [50%]\tLoss: -1.8704\n",
      "Training [60%]\tLoss: -2.2195\n",
      "Training [70%]\tLoss: -2.6295\n",
      "Training [80%]\tLoss: -3.1208\n",
      "Training [90%]\tLoss: -3.5538\n",
      "Training [100%]\tLoss: -4.1672\n"
     ]
    }
   ],
   "source": [
    "optimizer = optim.Adam(model4.parameters(), lr=0.001)\n",
    "loss_func = NLLLoss()\n",
    "\n",
    "\n",
    "epochs = 10     \n",
    "loss_list = []  \n",
    "model4.train()  \n",
    "for epoch in range(epochs):\n",
    "    total_loss = []\n",
    "    for batch_idx, (data, target) in enumerate(trainloader):\n",
    "        optimizer.zero_grad(set_to_none=True)  \n",
    "        output = model4(data)\n",
    "        loss = loss_func(output, target)      \n",
    "        loss.backward()                        \n",
    "        optimizer.step()                       \n",
    "        total_loss.append(loss.item())         \n",
    "    loss_list.append(sum(total_loss)/len(total_loss))\n",
    "    print('Training [{:.0f}%]\\tLoss: {:.4f}'.format(\n",
    "        100. * (epoch + 1) / epochs, loss_list[-1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zeHiMByWtvR_"
   },
   "source": [
    "**Plotting Loss:** Now, we need to visualize how our loss function behaved in each training iteration. We have plotted that in the following section:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 295
    },
    "id": "AJNwTqbMX7BU",
    "outputId": "8f6eb600-41d8-435d-b6bb-88cb1f764ca8"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(loss_list)\n",
    "plt.title('Hybrid NN Training Convergence')\n",
    "plt.xlabel('Training Iterations')\n",
    "plt.ylabel('Neg. Log Likelihood Loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gVSpSi3yuKSl"
   },
   "source": [
    "**Testing Hybrid QNN Model:** After training our Hybrid QNN model, we need to check the performance of our model on unknown datasets. In this case, we have our prepared test data. We have applied out model and found Accuracy: 83.4%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "iae8FverlKfb",
    "outputId": "8427fdaa-86fa-4530-8350-cc2346f43ec5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performance on test data:\n",
      "\tLoss: -4.2078\n",
      "\tAccuracy: 83.4%\n"
     ]
    }
   ],
   "source": [
    "model4.eval() \n",
    "with no_grad():\n",
    "\n",
    "    correct = 0\n",
    "    for batch_idx, (data, target) in enumerate(testloader):\n",
    "        output = model4(data)\n",
    "        if len(output.shape) == 1:\n",
    "            output = output.reshape(1, *output.shape)\n",
    "\n",
    "        pred = output.argmax(dim=1, keepdim=True)\n",
    "        correct += pred.eq(target.view_as(pred)).sum().item()\n",
    "\n",
    "        loss = loss_func(output, target)\n",
    "        total_loss.append(loss.item())\n",
    "\n",
    "    print('Performance on test data:\\n\\tLoss: {:.4f}\\n\\tAccuracy: {:.1f}%'\n",
    "          .format(sum(total_loss) / len(total_loss),\n",
    "                  correct / len(testloader) / batch_size * 100)\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EJFCyPUCuuLG"
   },
   "source": [
    "**Visualizing Results:** After performing our testing, we need to ensure our result visually. So, we have plotted our test data with the prediction done by our machine."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 130
    },
    "id": "r190sISVJU9A",
    "outputId": "8eea9524-0195-4909-9b6c-682891e07fef"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 6 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_samples_show = 6\n",
    "count = 0\n",
    "fig, axes = plt.subplots(nrows=1, ncols=n_samples_show, figsize=(10, 3))\n",
    "\n",
    "model4.eval()\n",
    "with no_grad():\n",
    "    for batch_idx, (data, target) in enumerate(testloader):\n",
    "        if count == n_samples_show:\n",
    "            break\n",
    "        output = model4(data[0:1])\n",
    "        if len(output.shape) == 1:\n",
    "            output = output.reshape(1, *output.shape)\n",
    "\n",
    "        pred = output.argmax(dim=1, keepdim=True)\n",
    "\n",
    "        axes[count].imshow(data[0].numpy().squeeze(), cmap='gray')\n",
    "\n",
    "        axes[count].set_xticks([])\n",
    "        axes[count].set_yticks([])\n",
    "        axes[count].set_title('Predicted {}'.format(pred.item()))\n",
    "\n",
    "        count += 1"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Image Classification using Qiskit_approach_2_Exp2.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}