E2E_SPC.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "authorship_tag": "ABX9TyN7vVUgjlWBS3KYKgVEQHoU",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/leonsuarez24/Notebooks/blob/main/E2E_SPC.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_tbTIVnjlIr1",
        "outputId": "e01d41e2-e51b-471d-9c04-f6e24931b1c8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting torchinfo\n",
            "  Downloading torchinfo-1.8.0-py3-none-any.whl (23 kB)\n",
            "Installing collected packages: torchinfo\n",
            "Successfully installed torchinfo-1.8.0\n"
          ]
        }
      ],
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms\n",
        "!pip install torchinfo\n",
        "from torchinfo import summary\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import torch.nn.functional as F\n",
        "\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5,), (0.5,))])\n",
        "trainset = torchvision.datasets.FashionMNIST(root='./data', train=True, download=True, transform=transform)\n",
        "trainset.data = transforms.Resize((32,32))(trainset.data)\n",
        "testset = torchvision.datasets.FashionMNIST(root='./data', train=False, download=True, transform=transform)\n",
        "testset.data = transforms.Resize((32,32))(testset.data)\n",
        "\n",
        "trainloader = torch.utils.data.DataLoader(trainset, batch_size=32, shuffle=True, num_workers=2)\n",
        "testloader = torch.utils.data.DataLoader(testset, batch_size=32, shuffle=False, num_workers=2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CyP3CLI_sG65",
        "outputId": "53432f01-6fd5-4b03-cb9e-1fbe9bc8ace7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz to ./data/FashionMNIST/raw/train-images-idx3-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 26421880/26421880 [00:01<00:00, 15833514.59it/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/FashionMNIST/raw/train-images-idx3-ubyte.gz to ./data/FashionMNIST/raw\n",
            "\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz to ./data/FashionMNIST/raw/train-labels-idx1-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 29515/29515 [00:00<00:00, 265845.94it/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/FashionMNIST/raw/train-labels-idx1-ubyte.gz to ./data/FashionMNIST/raw\n",
            "\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz to ./data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 4422102/4422102 [00:00<00:00, 5029648.90it/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz to ./data/FashionMNIST/raw\n",
            "\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz to ./data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 5148/5148 [00:00<00:00, 19736999.08it/s]\n",
            "/usr/local/lib/python3.10/dist-packages/torchvision/transforms/functional.py:1603: UserWarning: The default value of the antialias parameter of all the resizing transforms (Resize(), RandomResizedCrop(), etc.) will change from None to True in v0.17, in order to be consistent across the PIL and Tensor backends. To suppress this warning, directly pass antialias=True (recommended, future default), antialias=None (current default, which means False for Tensors and True for PIL), or antialias=False (only works on Tensors - PIL will still use antialiasing). This also applies if you are using the inference transforms from the models weights: update the call to weights.transforms(antialias=True).\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz to ./data/FashionMNIST/raw\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fig, ax = plt.subplots(5, 5)\n",
        "fig.tight_layout()\n",
        "fig.set_size_inches(10,10)\n",
        "k = 0\n",
        "for i in range(5):\n",
        "  for j in range(5):\n",
        "    ax[i, j].imshow(trainset.data[k], cmap='inferno')\n",
        "    ax[i, j].set_title(f\"{trainset.targets[k]}\")\n",
        "    ax[i, j].axis('off')\n",
        "    k +=1\n",
        "\n",
        "plt.show()\n",
        "print(trainset.data[0].shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 952
        },
        "id": "jsoR7UcTmIYW",
        "outputId": "6b0fdf86-918d-4523-a946-0893baae8d39"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x1000 with 25 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([32, 32])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ca = (torch.normal(0, 1, (10, 28, 28))/ np.sqrt(28*28))\n",
        "plt.imshow(ca[0,:,:])\n",
        "plt.colorbar()\n",
        "plt.show()\n",
        "print(torch.std(ca[0]), torch.mean(ca[0]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 447
        },
        "id": "4r-HJhHaKhNg",
        "outputId": "195a16fd-a367-4c3c-c990-2cc9e6b994cb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor(0.0350) tensor(-3.3425e-05)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "class OpticalLayer(nn.Module):\n",
        "    def __init__(self, K, output_dim, input_dim):\n",
        "        super(OpticalLayer, self).__init__()\n",
        "        self.output_dim = output_dim\n",
        "        self.input_dim = input_dim\n",
        "        self.K = K\n",
        "        ca = torch.normal(0, 1, (self.K, self.output_dim, self.input_dim))\n",
        "        self.weights = nn.Parameter(ca / np.sqrt(self.output_dim * self.input_dim))\n",
        "        #torch.nn.init.normal_(self.weights, mean=0.0, std=0.04)\n",
        "\n",
        "    def forward(self, x):\n",
        "        # Downscale to 1 px\n",
        "        ca_w = torch.unsqueeze(self.weights, 0)\n",
        "        x = torch.mul(x, ca_w)\n",
        "        x = torch.sum(x, dim=(-2, -1))\n",
        "        x = torch.unsqueeze(x, -1)\n",
        "        x = torch.unsqueeze(x, -1)\n",
        "        # Upscale to 28x28\n",
        "        x = torch.mul(x, ca_w)\n",
        "        x = torch.sum(x, dim=1)\n",
        "        x = torch.unsqueeze(x, 1)\n",
        "        x = x / torch.max(x)\n",
        "        return x"
      ],
      "metadata": {
        "id": "ClCflvHtx_Gu"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ol = OpticalLayer(160,32,32).to(device)\n",
        "summary(ol, input_size=(321, 1, 32, 32))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_RjXCT8bHXJ6",
        "outputId": "62fc5685-e0ca-4300-dcab-ae92ddebf679"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "==========================================================================================\n",
              "Layer (type:depth-idx)                   Output Shape              Param #\n",
              "==========================================================================================\n",
              "OpticalLayer                             [321, 1, 32, 32]          163,840\n",
              "==========================================================================================\n",
              "Total params: 163,840\n",
              "Trainable params: 163,840\n",
              "Non-trainable params: 0\n",
              "Total mult-adds (M): 52.59\n",
              "==========================================================================================\n",
              "Input size (MB): 1.31\n",
              "Forward/backward pass size (MB): 2.63\n",
              "Params size (MB): 0.66\n",
              "Estimated Total Size (MB): 4.60\n",
              "=========================================================================================="
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "class UNet(nn.Module):\n",
        "    def __init__(self, n_class):\n",
        "        super(UNet, self).__init__()\n",
        "\n",
        "        # Encoder\n",
        "        self.e11 = nn.Conv2d(1, 16, kernel_size=3, padding=1)  # Input: 1x32x32, Output: 16x32x32\n",
        "        self.e12 = nn.Conv2d(16, 16, kernel_size=3, padding=1)  # Output: 16x32x32\n",
        "        self.pool1 = nn.MaxPool2d(kernel_size=2, stride=2)  # Output: 16x16x16\n",
        "\n",
        "        self.e21 = nn.Conv2d(16, 32, kernel_size=3, padding=1)  # Output: 32x16x16\n",
        "        self.e22 = nn.Conv2d(32, 32, kernel_size=3, padding=1)  # Output: 32x16x16\n",
        "        self.pool2 = nn.MaxPool2d(kernel_size=2, stride=2)  # Output: 32x8x8\n",
        "\n",
        "        self.e31 = nn.Conv2d(32, 64, kernel_size=3, padding=1)  # Output: 64x8x8\n",
        "        self.e32 = nn.Conv2d(64, 64, kernel_size=3, padding=1)  # Output: 64x8x8\n",
        "        self.pool3 = nn.MaxPool2d(kernel_size=2, stride=2)  # Output: 64x4x4\n",
        "\n",
        "        self.e41 = nn.Conv2d(64, 128, kernel_size=3, padding=1)  # Output: 128x4x4\n",
        "        self.e42 = nn.Conv2d(128, 128, kernel_size=3, padding=1)  # Output: 128x4x4\n",
        "        self.pool4 = nn.MaxPool2d(kernel_size=2, stride=2)  # Output: 128x2x2\n",
        "\n",
        "        self.e51 = nn.Conv2d(128, 256, kernel_size=3, padding=1)  # Output: 256x2x2\n",
        "        self.e52 = nn.Conv2d(256, 256, kernel_size=3, padding=1)  # Output: 256x2x2\n",
        "\n",
        "\n",
        "        # Decoder\n",
        "        self.upconv1 = nn.ConvTranspose2d(256, 128, kernel_size=4, stride=2, padding=1)  # Output: 128x4x4\n",
        "        self.d11 = nn.Conv2d(256, 128, kernel_size=3, padding=1)\n",
        "        self.d12 = nn.Conv2d(128, 128, kernel_size=3, padding=1)\n",
        "\n",
        "        self.upconv2 = nn.ConvTranspose2d(128, 64, kernel_size=4, stride=2, padding=1)  # Output: 64x8x8\n",
        "        self.d21 = nn.Conv2d(128, 64, kernel_size=3, padding=1)\n",
        "        self.d22 = nn.Conv2d(64, 64, kernel_size=3, padding=1)\n",
        "\n",
        "        self.upconv3 = nn.ConvTranspose2d(64, 32, kernel_size=4, stride=2, padding=1)  # Output: 32x16x16\n",
        "        self.d31 = nn.Conv2d(64, 32, kernel_size=3, padding=1)\n",
        "        self.d32 = nn.Conv2d(32, 32, kernel_size=3, padding=1)\n",
        "\n",
        "        self.upconv4 = nn.ConvTranspose2d(32, 16, kernel_size=4, stride=2, padding=1)  # Output: 16x32x32\n",
        "        self.d41 = nn.Conv2d(32, 16, kernel_size=3, padding=1)\n",
        "        self.d42 = nn.Conv2d(16, 16, kernel_size=3, padding=1)\n",
        "\n",
        "        # Output layer\n",
        "        self.outconv = nn.Conv2d(16, n_class, kernel_size=1)\n",
        "\n",
        "    def forward(self, x):\n",
        "        # Encoder\n",
        "        xe11 = F.relu(self.e11(x))\n",
        "        xe12 = F.relu(self.e12(xe11))\n",
        "        xp1 = self.pool1(xe12)\n",
        "\n",
        "        xe21 = F.relu(self.e21(xp1))\n",
        "        xe22 = F.relu(self.e22(xe21))\n",
        "        xp2 = self.pool2(xe22)\n",
        "\n",
        "        xe31 = F.relu(self.e31(xp2))\n",
        "        xe32 = F.relu(self.e32(xe31))\n",
        "        xp3 = self.pool3(xe32)\n",
        "\n",
        "        xe41 = F.relu(self.e41(xp3))\n",
        "        xe42 = F.relu(self.e42(xe41))\n",
        "        xp4 = self.pool4(xe42)\n",
        "\n",
        "        xe51 = F.relu(self.e51(xp4))\n",
        "        xe52 = F.relu(self.e52(xe51))\n",
        "\n",
        "        # Decoder\n",
        "        xu1 = F.relu(self.upconv1(xe52))\n",
        "        xu11 = torch.cat([xu1, xe42], dim=1)\n",
        "        xd11 = F.relu(self.d11(xu11))\n",
        "        xd12 = F.relu(self.d12(xd11))\n",
        "\n",
        "        xu2 = F.relu(self.upconv2(xd12))\n",
        "        xu22 = torch.cat([xu2, xe32], dim=1)\n",
        "        xd21 = F.relu(self.d21(xu22))\n",
        "        xd22 = F.relu(self.d22(xd21))\n",
        "\n",
        "        xu3 = F.relu(self.upconv3(xd22))\n",
        "        xu33 = torch.cat([xu3, xe22], dim=1)\n",
        "        xd31 = F.relu(self.d31(xu33))\n",
        "        xd32 = F.relu(self.d32(xd31))\n",
        "\n",
        "        xu4 = F.relu(self.upconv4(xd32))\n",
        "        xu44 = torch.cat([xu4, xe12], dim=1)\n",
        "        xd41 = F.relu(self.d41(xu44))\n",
        "        xd42 = F.relu(self.d42(xd41))\n",
        "\n",
        "        # Output layer\n",
        "        out = self.outconv(xd42)\n",
        "\n",
        "        return out"
      ],
      "metadata": {
        "id": "8Gj26jkdPiG3"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "unet = UNet(1)\n",
        "print(summary(unet, input_size=(12, 1, 32, 32)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "n0r_mQZwccIH",
        "outputId": "abcf2cbc-57a5-4651-e024-f4589867259e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "==========================================================================================\n",
            "Layer (type:depth-idx)                   Output Shape              Param #\n",
            "==========================================================================================\n",
            "UNet                                     [12, 1, 32, 32]           --\n",
            "├─Conv2d: 1-1                            [12, 16, 32, 32]          160\n",
            "├─Conv2d: 1-2                            [12, 16, 32, 32]          2,320\n",
            "├─MaxPool2d: 1-3                         [12, 16, 16, 16]          --\n",
            "├─Conv2d: 1-4                            [12, 32, 16, 16]          4,640\n",
            "├─Conv2d: 1-5                            [12, 32, 16, 16]          9,248\n",
            "├─MaxPool2d: 1-6                         [12, 32, 8, 8]            --\n",
            "├─Conv2d: 1-7                            [12, 64, 8, 8]            18,496\n",
            "├─Conv2d: 1-8                            [12, 64, 8, 8]            36,928\n",
            "├─MaxPool2d: 1-9                         [12, 64, 4, 4]            --\n",
            "├─Conv2d: 1-10                           [12, 128, 4, 4]           73,856\n",
            "├─Conv2d: 1-11                           [12, 128, 4, 4]           147,584\n",
            "├─MaxPool2d: 1-12                        [12, 128, 2, 2]           --\n",
            "├─Conv2d: 1-13                           [12, 256, 2, 2]           295,168\n",
            "├─Conv2d: 1-14                           [12, 256, 2, 2]           590,080\n",
            "├─ConvTranspose2d: 1-15                  [12, 128, 4, 4]           524,416\n",
            "├─Conv2d: 1-16                           [12, 128, 4, 4]           295,040\n",
            "├─Conv2d: 1-17                           [12, 128, 4, 4]           147,584\n",
            "├─ConvTranspose2d: 1-18                  [12, 64, 8, 8]            131,136\n",
            "├─Conv2d: 1-19                           [12, 64, 8, 8]            73,792\n",
            "├─Conv2d: 1-20                           [12, 64, 8, 8]            36,928\n",
            "├─ConvTranspose2d: 1-21                  [12, 32, 16, 16]          32,800\n",
            "├─Conv2d: 1-22                           [12, 32, 16, 16]          18,464\n",
            "├─Conv2d: 1-23                           [12, 32, 16, 16]          9,248\n",
            "├─ConvTranspose2d: 1-24                  [12, 16, 32, 32]          8,208\n",
            "├─Conv2d: 1-25                           [12, 16, 32, 32]          4,624\n",
            "├─Conv2d: 1-26                           [12, 16, 32, 32]          2,320\n",
            "├─Conv2d: 1-27                           [12, 1, 32, 32]           17\n",
            "==========================================================================================\n",
            "Total params: 2,463,057\n",
            "Trainable params: 2,463,057\n",
            "Non-trainable params: 0\n",
            "Total mult-adds (M): 944.42\n",
            "==========================================================================================\n",
            "Input size (MB): 0.05\n",
            "Forward/backward pass size (MB): 15.04\n",
            "Params size (MB): 9.85\n",
            "Estimated Total Size (MB): 24.94\n",
            "==========================================================================================\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "class E2E(nn.Module):\n",
        "  def __init__(self, K, output_dim, input_dim, n_class):\n",
        "    super(E2E, self).__init__()\n",
        "    self.optical_layer = OpticalLayer(K, output_dim, input_dim)\n",
        "    self.UNET = UNet(n_class)\n",
        "\n",
        "  def forward(self, x):\n",
        "    out = self.optical_layer(x)\n",
        "    out = self.UNET(out)\n",
        "    return out"
      ],
      "metadata": {
        "id": "i5IpB3qPckwP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "e2e = E2E(160, 32, 32, 1)\n",
        "print(summary(e2e, input_size=(12, 1, 32, 32)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IUp1Bsd-rM8F",
        "outputId": "0fc0b9db-bb69-482f-cd76-5962217b933f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "==========================================================================================\n",
            "Layer (type:depth-idx)                   Output Shape              Param #\n",
            "==========================================================================================\n",
            "E2E                                      [12, 1, 32, 32]           --\n",
            "├─OpticalLayer: 1-1                      [12, 1, 32, 32]           163,840\n",
            "├─UNet: 1-2                              [12, 1, 32, 32]           --\n",
            "│    └─Conv2d: 2-1                       [12, 16, 32, 32]          160\n",
            "│    └─Conv2d: 2-2                       [12, 16, 32, 32]          2,320\n",
            "│    └─MaxPool2d: 2-3                    [12, 16, 16, 16]          --\n",
            "│    └─Conv2d: 2-4                       [12, 32, 16, 16]          4,640\n",
            "│    └─Conv2d: 2-5                       [12, 32, 16, 16]          9,248\n",
            "│    └─MaxPool2d: 2-6                    [12, 32, 8, 8]            --\n",
            "│    └─Conv2d: 2-7                       [12, 64, 8, 8]            18,496\n",
            "│    └─Conv2d: 2-8                       [12, 64, 8, 8]            36,928\n",
            "│    └─MaxPool2d: 2-9                    [12, 64, 4, 4]            --\n",
            "│    └─Conv2d: 2-10                      [12, 128, 4, 4]           73,856\n",
            "│    └─Conv2d: 2-11                      [12, 128, 4, 4]           147,584\n",
            "│    └─MaxPool2d: 2-12                   [12, 128, 2, 2]           --\n",
            "│    └─Conv2d: 2-13                      [12, 256, 2, 2]           295,168\n",
            "│    └─Conv2d: 2-14                      [12, 256, 2, 2]           590,080\n",
            "│    └─ConvTranspose2d: 2-15             [12, 128, 4, 4]           524,416\n",
            "│    └─Conv2d: 2-16                      [12, 128, 4, 4]           295,040\n",
            "│    └─Conv2d: 2-17                      [12, 128, 4, 4]           147,584\n",
            "│    └─ConvTranspose2d: 2-18             [12, 64, 8, 8]            131,136\n",
            "│    └─Conv2d: 2-19                      [12, 64, 8, 8]            73,792\n",
            "│    └─Conv2d: 2-20                      [12, 64, 8, 8]            36,928\n",
            "│    └─ConvTranspose2d: 2-21             [12, 32, 16, 16]          32,800\n",
            "│    └─Conv2d: 2-22                      [12, 32, 16, 16]          18,464\n",
            "│    └─Conv2d: 2-23                      [12, 32, 16, 16]          9,248\n",
            "│    └─ConvTranspose2d: 2-24             [12, 16, 32, 32]          8,208\n",
            "│    └─Conv2d: 2-25                      [12, 16, 32, 32]          4,624\n",
            "│    └─Conv2d: 2-26                      [12, 16, 32, 32]          2,320\n",
            "│    └─Conv2d: 2-27                      [12, 1, 32, 32]           17\n",
            "==========================================================================================\n",
            "Total params: 2,626,897\n",
            "Trainable params: 2,626,897\n",
            "Non-trainable params: 0\n",
            "Total mult-adds (M): 946.38\n",
            "==========================================================================================\n",
            "Input size (MB): 0.05\n",
            "Forward/backward pass size (MB): 15.14\n",
            "Params size (MB): 10.51\n",
            "Estimated Total Size (MB): 25.70\n",
            "==========================================================================================\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "optimizer = torch.optim.Adam(e2e.parameters())\n",
        "criterion = torch.nn.MSELoss()\n",
        "epochs = 10\n",
        "regularizer_parameter = 1e-3\n",
        "\n",
        "for epoch in range(epochs):\n",
        "  running_loss = 0.0\n",
        "  for i, data in enumerate(trainloader, 0):\n",
        "    inputs, _ = data\n",
        "    inputs = inputs.to(device)\n",
        "    optimizer.zero_grad()\n",
        "\n",
        "    pred = e2e(inputs)\n",
        "    loss = criterion(pred, inputs) + regularizer_parameter * (torch.sum(torch.mul(torch.pow(1 + e2e.optical_layer.weights, 2), torch.pow(1 - e2e.optical_layer.weights, 2))))\n",
        "\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "\n",
        "    running_loss += loss.item()\n",
        "    if (i + 1) % 375 == 0:  # Print the running loss every 375 mini-batches\n",
        "      print(f'Epoch: {epoch + 1}, Batch: {i + 1}, Loss: {running_loss / 375:.3f}')\n",
        "      running_loss = 0.0\n",
        "\n",
        "print('Training finished!')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "x6uK_dHByTt0",
        "outputId": "17873303-d049-415c-c6b2-065189855250"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 1, Batch: 375, Loss: 153.379\n",
            "Epoch: 1, Batch: 750, Loss: 65.950\n",
            "Epoch: 1, Batch: 1125, Loss: 6.261\n",
            "Epoch: 1, Batch: 1500, Loss: 1.194\n",
            "Epoch: 1, Batch: 1875, Loss: 0.659\n",
            "Epoch: 2, Batch: 375, Loss: 0.404\n",
            "Epoch: 2, Batch: 750, Loss: 0.265\n",
            "Epoch: 2, Batch: 1125, Loss: 0.191\n",
            "Epoch: 2, Batch: 1500, Loss: 0.152\n",
            "Epoch: 2, Batch: 1875, Loss: 0.123\n",
            "Epoch: 3, Batch: 375, Loss: 0.097\n",
            "Epoch: 3, Batch: 750, Loss: 0.076\n",
            "Epoch: 3, Batch: 1125, Loss: 0.059\n",
            "Epoch: 3, Batch: 1500, Loss: 0.045\n",
            "Epoch: 3, Batch: 1875, Loss: 0.034\n",
            "Epoch: 4, Batch: 375, Loss: 0.027\n",
            "Epoch: 4, Batch: 750, Loss: 0.023\n",
            "Epoch: 4, Batch: 1125, Loss: 0.021\n",
            "Epoch: 4, Batch: 1500, Loss: 0.021\n",
            "Epoch: 4, Batch: 1875, Loss: 0.020\n",
            "Epoch: 5, Batch: 375, Loss: 0.020\n",
            "Epoch: 5, Batch: 750, Loss: 0.020\n",
            "Epoch: 5, Batch: 1125, Loss: 0.020\n",
            "Epoch: 5, Batch: 1500, Loss: 0.020\n",
            "Epoch: 5, Batch: 1875, Loss: 0.019\n",
            "Epoch: 6, Batch: 375, Loss: 0.019\n",
            "Epoch: 6, Batch: 750, Loss: 0.019\n",
            "Epoch: 6, Batch: 1125, Loss: 0.019\n",
            "Epoch: 6, Batch: 1500, Loss: 0.019\n",
            "Epoch: 6, Batch: 1875, Loss: 0.019\n",
            "Epoch: 7, Batch: 375, Loss: 0.018\n",
            "Epoch: 7, Batch: 750, Loss: 0.018\n",
            "Epoch: 7, Batch: 1125, Loss: 0.018\n",
            "Epoch: 7, Batch: 1500, Loss: 0.018\n",
            "Epoch: 7, Batch: 1875, Loss: 0.018\n",
            "Epoch: 8, Batch: 375, Loss: 0.018\n",
            "Epoch: 8, Batch: 750, Loss: 0.017\n",
            "Epoch: 8, Batch: 1125, Loss: 0.018\n",
            "Epoch: 8, Batch: 1500, Loss: 0.018\n",
            "Epoch: 8, Batch: 1875, Loss: 0.018\n",
            "Epoch: 9, Batch: 375, Loss: 0.017\n",
            "Epoch: 9, Batch: 750, Loss: 0.017\n",
            "Epoch: 9, Batch: 1125, Loss: 0.017\n",
            "Epoch: 9, Batch: 1500, Loss: 0.017\n",
            "Epoch: 9, Batch: 1875, Loss: 0.017\n",
            "Epoch: 10, Batch: 375, Loss: 0.017\n",
            "Epoch: 10, Batch: 750, Loss: 0.017\n",
            "Epoch: 10, Batch: 1125, Loss: 0.017\n",
            "Epoch: 10, Batch: 1500, Loss: 0.017\n",
            "Epoch: 10, Batch: 1875, Loss: 0.017\n",
            "Training finished!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "e2e.optical_layer.weights.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2SagDbwq4uSS",
        "outputId": "4644849f-8d9f-47d7-d1c7-92aa4d3dd831"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([160, 32, 32])"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_images(images):\n",
        "    fig, axes = plt.subplots(5, 5, figsize=(10, 10))\n",
        "    axes = axes.ravel()\n",
        "    for i in range(len(images)):\n",
        "        axes[i].imshow(images[i].squeeze(), cmap='gray')\n",
        "        axes[i].set_title(f'CA: {i + 1}')\n",
        "        axes[i].axis('off')\n",
        "    plt.tight_layout()\n",
        "    plt.show()"
      ],
      "metadata": {
        "id": "_q_qdjOA6ost"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "subset_indices = torch.randint(0, len(testset), (25,))\n",
        "subset_images = [testset[i][0] for i in subset_indices]\n",
        "subset_labels = [testset[i][1] for i in subset_indices]\n",
        "subset_images = torch.stack(subset_images).to(device)\n",
        "e2e.eval()\n",
        "spc = e2e.cpu()\n",
        "subset_images = subset_images.cpu()\n",
        "\n",
        "with torch.no_grad():\n",
        "    outputs = spc(subset_images)\n",
        "    _, predicted = torch.max(outputs, 1)\n",
        "    print(e2e.optical_layer.weights.shape)\n",
        "    plot_images(e2e.optical_layer.weights[:25,:,:])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "_5gMiDYX6dQT",
        "outputId": "f95a6cb7-168e-41ee-c64e-9420d6dce6ba"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([160, 32, 32])\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x1000 with 25 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluation\n",
        "reconstruced_imgs=[]\n",
        "with torch.no_grad():\n",
        "  for data in testloader:\n",
        "    images, _ = data\n",
        "    images = images.to(\"cpu\")\n",
        "    reconstruced_imgs.append(torch.squeeze(e2e(images), dim = 1))"
      ],
      "metadata": {
        "id": "lcLycR26ARA6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(reconstruced_imgs[0].shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2JX6tmXQCXrC",
        "outputId": "c3020f6e-9a1e-4df6-f260-ddfeff7bc8f3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([32, 32, 32])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "index = 9000\n",
        "batch = index//32\n",
        "f, ax = plt.subplots(1,2)\n",
        "f.set_figheight(10)\n",
        "f.set_figwidth(10)\n",
        "ax[0].imshow(testset.data[index],cmap='inferno')\n",
        "ax[1].imshow(torch.squeeze(reconstruced_imgs[batch][index - batch*32]),cmap='inferno')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 412
        },
        "id": "u2YVFwSQBIt3",
        "outputId": "07d28cbe-284f-4b5e-f4dd-25b76e298933"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x1000 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "index = 9500\n",
        "batch = index//32\n",
        "f, ax = plt.subplots(1,2)\n",
        "f.set_figheight(10)\n",
        "f.set_figwidth(10)\n",
        "ax[0].imshow(testset.data[index],cmap='inferno')\n",
        "ax[1].imshow(torch.squeeze(reconstruced_imgs[batch][index - batch*32]),cmap='inferno')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 412
        },
        "id": "JfCXsPinBm6u",
        "outputId": "35ec98cd-7359-44e1-a01a-3bc8f0ae7a68"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x1000 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "5fraTqh3FIVs"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}