Hybrid NN FashionMNIST.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "kEkU8Y9I8xt3"
      },
      "outputs": [],
      "source": [
        "# !pip install -r 'requirements.txt' --quiet"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "ncmziV32-sdn"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "\n",
        "import torch\n",
        "from torchvision.transforms import ToTensor\n",
        "from torch import no_grad\n",
        "from torchvision import datasets\n",
        "import torch.optim as optim\n",
        "from torch.nn import (Module, Conv2d, Linear, Dropout2d, NLLLoss, MaxPool2d, Flatten, Sequential, ReLU)\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "from torch.utils.data import DataLoader\n",
        "\n",
        "import qiskit\n",
        "from qiskit import transpile, assemble\n",
        "from qiskit.visualization import *\n",
        "from qiskit  import Aer, QuantumCircuit\n",
        "from qiskit.utils import QuantumInstance\n",
        "from qiskit.opflow import AerPauliExpectation\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\n",
        "import matplotlib.pyplot as plt"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "Xd20_u7I-xOA"
      },
      "outputs": [],
      "source": [
        "# Declare Quantum instance\n",
        "qi = QuantumInstance(Aer.get_backend('aer_simulator_statevector'))\n",
        "\n",
        "### Training and test data downloaded from FashionMNIST and transformed into tensors ###\n",
        "training_data = datasets.FashionMNIST(\n",
        "    root=\"data\",\n",
        "    train=True,\n",
        "    download=True,\n",
        "    transform=ToTensor()\n",
        ")\n",
        "\n",
        "test_data = datasets.FashionMNIST(\n",
        "    root=\"data\",\n",
        "    train=False,\n",
        "    download=True,\n",
        "    transform=ToTensor()\n",
        ")\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 482
        },
        "id": "fM67SnqC-1qU",
        "outputId": "1b834eae-2e9a-431d-8984-ed79d94c91e3"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 576x576 with 9 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "### Inspecting the images in the training data set with their labels ###\n",
        "labels_map = {\n",
        "    0: \"T-Shirt\",\n",
        "    1: \"Trouser\",\n",
        "    2: \"Pullover\",\n",
        "    3: \"Dress\",\n",
        "    4: \"Coat\",\n",
        "    5: \"Sandal\",\n",
        "    6: \"Shirt\",\n",
        "    7: \"Sneaker\",\n",
        "    8: \"Bag\",\n",
        "    9: \"Ankle Boot\",\n",
        "}\n",
        "\n",
        "figure = plt.figure(figsize=(8, 8))\n",
        "cols, rows = 3, 3\n",
        "for i in range(1, cols * rows + 1):\n",
        "    sample_idx = torch.randint(len(training_data), size=(1,)).item()\n",
        "    img, label = training_data[sample_idx]\n",
        "    figure.add_subplot(rows, cols, i)\n",
        "    plt.title(labels_map[label])\n",
        "    plt.axis(\"off\")\n",
        "    plt.imshow(img.squeeze(), cmap=\"gray\")\n",
        "plt.show()\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "OiZpcaW8-6As"
      },
      "outputs": [],
      "source": [
        "### Load training data into Torch DataLoader ###\n",
        "X_train = training_data\n",
        "n_samples = 500\n",
        "batch_size = 64\n",
        "\n",
        "# Filter out labels (originally 0-9), leaving only labels 0 and 1\n",
        "idx = np.append(np.where(X_train.targets == 0)[0][:n_samples],\n",
        "                np.where(X_train.targets == 1)[0][:n_samples])\n",
        "\n",
        "X_train.data = X_train.data[idx]\n",
        "X_train.targets = X_train.targets[idx]\n",
        "\n",
        "# A torch dataloader is defined with filtered data\n",
        "train_loader = DataLoader(X_train, batch_size=64, shuffle=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "JWDp-YBZ-8MO"
      },
      "outputs": [],
      "source": [
        "# Load test data into Torch DataLoader\n",
        "X_test = test_data\n",
        "\n",
        "# Filter out labels (originally 0-9), leaving only labels 0 and 1\n",
        "idx = np.append(np.where(X_test.targets == 0)[0][:n_samples],\n",
        "                np.where(X_test.targets == 1)[0][:n_samples])\n",
        "\n",
        "X_test.data = X_test.data[idx]\n",
        "X_test.targets = X_test.targets[idx]\n",
        "\n",
        "# Define torch dataloader with filtered data\n",
        "test_loader = DataLoader(X_test, batch_size=64, shuffle=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t58Ch9-0_Mjj",
        "outputId": "42b2ed62-d6cf-4cde-9d9d-7f794b06875c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "ComposedOp([\n",
            "  OperatorMeasurement(1.0 * ZZ),\n",
            "  CircuitStateFn(\n",
            "       ┌──────────────────────────┐┌──────────────────────────────────────┐\n",
            "  q_0: ┤0                         ├┤0                                     ├\n",
            "       │  ZZFeatureMap(x[0],x[1]) ││  RealAmplitudes(θ[0],θ[1],θ[2],θ[3]) │\n",
            "  q_1: ┤1                         ├┤1                                     ├\n",
            "       └──────────────────────────┘└──────────────────────────────────────┘\n",
            "  )\n",
            "])\n"
          ]
        }
      ],
      "source": [
        "### Two layer QNN constructed ###\n",
        "feature_map = ZZFeatureMap(feature_dimension=2, entanglement='linear')\n",
        "ansatz = RealAmplitudes(2, reps=1, entanglement='linear')\n",
        "qnn2 = TwoLayerQNN(2, feature_map, ansatz, input_gradients=True,\n",
        "                   exp_val=AerPauliExpectation(), quantum_instance=qi)\n",
        "print(qnn2.operator)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "R7TCFuw__OQL"
      },
      "outputs": [],
      "source": [
        "### Torch NN module from Qiskit ###\n",
        "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)         # 2-dimensional input to QNN\n",
        "        # Apply torch connector, weights chosen\n",
        "        self.qnn = TorchConnector(qnn2)\n",
        "        # uniformly at random from interval [-1,1].\n",
        "        self.fc3 = Linear(1, 1)\n",
        "        # 1-dimensional output from QNN\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(x.shape[0], -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 torch.cat((x, 1 - x), -1)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YwX5zShO_P2m",
        "outputId": "bfe7271b-bee6-4a2f-96dc-4b4277e6af08"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Training [5%]\tLoss: -0.5811\n",
            "Training [10%]\tLoss: -0.7120\n",
            "Training [15%]\tLoss: -0.7710\n",
            "Training [20%]\tLoss: -0.8145\n",
            "Training [25%]\tLoss: -0.8467\n",
            "Training [30%]\tLoss: -0.8669\n",
            "Training [35%]\tLoss: -0.9091\n",
            "Training [40%]\tLoss: -0.9372\n",
            "Training [45%]\tLoss: -0.9804\n",
            "Training [50%]\tLoss: -1.0021\n",
            "Training [55%]\tLoss: -1.0286\n",
            "Training [60%]\tLoss: -1.0652\n",
            "Training [65%]\tLoss: -1.0867\n",
            "Training [70%]\tLoss: -1.1223\n",
            "Training [75%]\tLoss: -1.1481\n",
            "Training [80%]\tLoss: -1.1706\n",
            "Training [85%]\tLoss: -1.1992\n",
            "Training [90%]\tLoss: -1.2255\n",
            "Training [95%]\tLoss: -1.2485\n",
            "Training [100%]\tLoss: -1.2707\n"
          ]
        }
      ],
      "source": [
        "### Model trained and the loss computed ###\n",
        "model = Net()\n",
        "optimizer = optim.Adam(model.parameters(), lr=0.001)\n",
        "loss_func = nn.NLLLoss()\n",
        "\n",
        "epochs = 20\n",
        "loss_list = []\n",
        "\n",
        "model.train()\n",
        "for epoch in range(epochs):\n",
        "    total_loss = []\n",
        "    for batch_idx, (data, target) in enumerate(train_loader):\n",
        "        optimizer.zero_grad()\n",
        "        # Forward pass\n",
        "        output = model(data)\n",
        "        # Calculating loss\n",
        "        loss = loss_func(output, target)\n",
        "        # Backward pass\n",
        "        loss.backward()\n",
        "        # Optimize the weights\n",
        "        optimizer.step()\n",
        "\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]))\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "gS2fX06B_RPn",
        "outputId": "90b9e918-f97f-4669-9a48-afe02cf1a5f3"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "### Loss convergence plotted ###\n",
        "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()\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_SGwgOk_CQPY",
        "outputId": "5491b63d-201d-49d8-fba8-fe1ea3c503bf"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Performance on test data:\n",
            "\tLoss: -1.2303\n",
            "\tAccuracy: 95.0%\n"
          ]
        }
      ],
      "source": [
        "### Model evaluated ###\n",
        "model.eval()\n",
        "with torch.no_grad():\n",
        "    \n",
        "    correct = 0\n",
        "    for batch_idx, (data, target) in enumerate(test_loader):\n",
        "        output = model(data)\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}%'.format(\n",
        "        sum(total_loss) / len(total_loss),\n",
        "        (correct / len(test_loader) / batch_size) * 100))\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        },
        "id": "4x4zpOV1CUWf",
        "outputId": "c2ff8978-896e-4bd3-8618-68baf177aa18"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1080x360 with 6 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "### Predicted images displayed. Either T-shirt or Trouser ###\n",
        "n_samples_show = 6\n",
        "count = 0\n",
        "fig, axes = plt.subplots(nrows=1, ncols=n_samples_show, figsize=(15, 5))\n",
        "\n",
        "model.eval()\n",
        "with no_grad():\n",
        "    for batch_idx, (data, target) in enumerate(test_loader):\n",
        "        if count == n_samples_show:\n",
        "            break\n",
        "        output = model(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",
        "        axes[count].imshow(data[0].numpy().squeeze(), cmap='gray')\n",
        "\n",
        "        axes[count].set_xticks([])\n",
        "        axes[count].set_yticks([])\n",
        "        if pred.item() == 0:\n",
        "            axes[count].set_title('Predicted item: T-Shirt')  \n",
        "        elif pred.item() == 1:\n",
        "            axes[count].set_title('Predicted item: Trouser')  \n",
        "        count += 1"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [],
      "name": "Untitled2.ipynb",
      "provenance": []
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "Python 3.8.12 ('qiskit-env')",
      "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": "3.8.12"
    },
    "vscode": {
      "interpreter": {
        "hash": "c5e597ca477812040ed5a1629c6451aa891ac46335823a5f6165f6c73dfc5551"
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}