CNN_torch.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "authorship_tag": "ABX9TyPLMJ3TmkL1vDfaNkCfxaLZ",
      "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/chellick/Algorithms_comparison/blob/main/CNN_torch.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": {
        "id": "s2VFwyPpGZy2"
      },
      "outputs": [],
      "source": [
        "import torch\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from tensorflow.keras.datasets import cifar10 #ignore\n",
        "import tensorflow as tf\n",
        "from torch import nn\n",
        "from torch.utils.data import DataLoader, TensorDataset\n",
        "import time"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "(x_train, _), (x_test, _) = cifar10.load_data()\n",
        "x_train = x_train.astype('float32') / 255.0\n",
        "x_test = x_test.astype('float32') / 255.0"
      ],
      "metadata": {
        "id": "txrYhE6i0u3A",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c02e9d26-30fe-4956-c48d-f2fb9dc3827a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n",
            "170498071/170498071 [==============================] - 14s 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_train_lr = torch.tensor(np.array(tf.image.resize(x_train, (x_train.shape[1] // 2, x_train.shape[2] // 2))), dtype=torch.float32)[:7000]\n",
        "x_test_lr = torch.tensor(np.array(tf.image.resize(x_test, (x_test.shape[1] // 2, x_test.shape[2] // 2))), dtype=torch.float32)[:3000]\n",
        "\n",
        "# Convert to float32 for x_train_hr and x_test_hr\n",
        "x_train_hr = torch.tensor(x_train, dtype=torch.float32)[:7000]\n",
        "x_test_hr = torch.tensor(x_test, dtype=torch.float32)[:3000]\n",
        "\n",
        "x_train_lr = x_train_lr.permute(0, 3, 1, 2)\n",
        "x_test_lr = x_test_lr.permute(0, 3, 1, 2)\n",
        "\n",
        "x_train_hr = x_train_hr.permute(0, 3, 1, 2)\n",
        "x_test_hr = x_test_hr.permute(0, 3, 1, 2)\n",
        "\n",
        "\n",
        "train_dataset = TensorDataset(x_train_lr, x_train_hr)\n",
        "train_loader = DataLoader(train_dataset, batch_size=64, shuffle=True)"
      ],
      "metadata": {
        "id": "R24ZRUDJJYVP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from torch.nn.modules.upsampling import Upsample\n",
        "class CNN(nn.Module):\n",
        "  def __init__(self):\n",
        "    super(CNN, self).__init__()\n",
        "    self.layers = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=3, out_channels=32, kernel_size=(3,3), padding=\"same\"),\n",
        "            nn.ReLU(),\n",
        "            nn.Conv2d(in_channels=32, out_channels=32, kernel_size=(3,3), padding=\"same\"),\n",
        "            nn.BatchNorm2d(32),\n",
        "            nn.ReLU(),\n",
        "            nn.Conv2d(in_channels=32, out_channels=3, kernel_size=(3,3), padding=\"same\"),\n",
        "            nn.BatchNorm2d(3),\n",
        "            nn.ReLU(),\n",
        "            nn.Conv2d(in_channels=3, out_channels=3, kernel_size=(3,3), padding=\"same\"),\n",
        "            nn.Upsample(size=(32, 32), mode='bilinear'),\n",
        "    )\n",
        "\n",
        "\n",
        "  def forward(self, x):\n",
        "    return self.layers(x)\n"
      ],
      "metadata": {
        "id": "7DdAhhBcJlZS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = CNN()\n",
        "loss_fn = nn.L1Loss()\n",
        "optimizer = torch.optim.Adam(model.parameters(), lr=0.0001)\n",
        "model.parameters"
      ],
      "metadata": {
        "id": "_X_nHIYiM9x5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "87831f07-da9c-433c-ab0f-d44868f6e91c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<bound method Module.parameters of CNN(\n",
              "  (layers): Sequential(\n",
              "    (0): Conv2d(3, 32, kernel_size=(3, 3), stride=(1, 1), padding=same)\n",
              "    (1): ReLU()\n",
              "    (2): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=same)\n",
              "    (3): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "    (4): ReLU()\n",
              "    (5): Conv2d(32, 3, kernel_size=(3, 3), stride=(1, 1), padding=same)\n",
              "    (6): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
              "    (7): ReLU()\n",
              "    (8): Conv2d(3, 3, kernel_size=(3, 3), stride=(1, 1), padding=same)\n",
              "    (9): Upsample(size=(32, 32), mode='bilinear')\n",
              "  )\n",
              ")>"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction = model(x_train_lr[:1])\n",
        "prediction = prediction.permute(0, 2, 3, 1)\n",
        "prediction.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SFyNLIZvRIH7",
        "outputId": "9d015e87-f45d-4626-cfb0-f437748ae3e5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([1, 32, 32, 3])"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.imshow(prediction[0].detach().numpy())\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 467
        },
        "id": "VsKkKkbUX3Fi",
        "outputId": "05cb4cb5-cb3c-4591-879b-54bb491a7eb7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for epoch in range(100):\n",
        "    model.train()\n",
        "    start_time = time.time()\n",
        "    for batch_x, batch_y in train_loader:\n",
        "        pred = model(batch_x)\n",
        "        loss = loss_fn(pred, batch_y)\n",
        "        optimizer.zero_grad()\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "\n",
        "    print(\" Loss:\", loss.item(), \"\\n\", \"Epoch:\", epoch, \"\\n\", \"Time:\",  time.time() - start_time, \"\\n\")\n"
      ],
      "metadata": {
        "id": "mXqSW7X3M2Y5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "83ef3826-769d-48d6-84ec-7d2fbc98ef2d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            " Loss: 0.2969021797180176 \n",
            " Epoch: 0 \n",
            " Time: 4.763331174850464 \n",
            "\n",
            " Loss: 0.25435441732406616 \n",
            " Epoch: 1 \n",
            " Time: 3.736159324645996 \n",
            "\n",
            " Loss: 0.14900167286396027 \n",
            " Epoch: 2 \n",
            " Time: 3.7872073650360107 \n",
            "\n",
            " Loss: 0.13812586665153503 \n",
            " Epoch: 3 \n",
            " Time: 4.662527084350586 \n",
            "\n",
            " Loss: 0.09852062165737152 \n",
            " Epoch: 4 \n",
            " Time: 3.9190611839294434 \n",
            "\n",
            " Loss: 0.08666941523551941 \n",
            " Epoch: 5 \n",
            " Time: 3.819526195526123 \n",
            "\n",
            " Loss: 0.08888270705938339 \n",
            " Epoch: 6 \n",
            " Time: 4.5220184326171875 \n",
            "\n",
            " Loss: 0.08469505608081818 \n",
            " Epoch: 7 \n",
            " Time: 3.909118890762329 \n",
            "\n",
            " Loss: 0.07463256269693375 \n",
            " Epoch: 8 \n",
            " Time: 3.781724214553833 \n",
            "\n",
            " Loss: 0.08387655019760132 \n",
            " Epoch: 9 \n",
            " Time: 4.244129657745361 \n",
            "\n",
            " Loss: 0.07469023764133453 \n",
            " Epoch: 10 \n",
            " Time: 4.2007646560668945 \n",
            "\n",
            " Loss: 0.0769849419593811 \n",
            " Epoch: 11 \n",
            " Time: 3.794578790664673 \n",
            "\n",
            " Loss: 0.06881865859031677 \n",
            " Epoch: 12 \n",
            " Time: 4.201437473297119 \n",
            "\n",
            " Loss: 0.06745510548353195 \n",
            " Epoch: 13 \n",
            " Time: 4.563258409500122 \n",
            "\n",
            " Loss: 0.06141873076558113 \n",
            " Epoch: 14 \n",
            " Time: 4.17953085899353 \n",
            "\n",
            " Loss: 0.07337471842765808 \n",
            " Epoch: 15 \n",
            " Time: 4.619033575057983 \n",
            "\n",
            " Loss: 0.062438759952783585 \n",
            " Epoch: 16 \n",
            " Time: 3.9198272228240967 \n",
            "\n",
            " Loss: 0.05974040925502777 \n",
            " Epoch: 17 \n",
            " Time: 3.707094192504883 \n",
            "\n",
            " Loss: 0.06445678323507309 \n",
            " Epoch: 18 \n",
            " Time: 4.439330577850342 \n",
            "\n",
            " Loss: 0.06789024919271469 \n",
            " Epoch: 19 \n",
            " Time: 4.123634099960327 \n",
            "\n",
            " Loss: 0.062443096190690994 \n",
            " Epoch: 20 \n",
            " Time: 3.7510578632354736 \n",
            "\n",
            " Loss: 0.05948315188288689 \n",
            " Epoch: 21 \n",
            " Time: 4.414841890335083 \n",
            "\n",
            " Loss: 0.06293988227844238 \n",
            " Epoch: 22 \n",
            " Time: 4.0988686084747314 \n",
            "\n",
            " Loss: 0.06014801561832428 \n",
            " Epoch: 23 \n",
            " Time: 3.7518882751464844 \n",
            "\n",
            " Loss: 0.06315652281045914 \n",
            " Epoch: 24 \n",
            " Time: 4.23650860786438 \n",
            "\n",
            " Loss: 0.05827416852116585 \n",
            " Epoch: 25 \n",
            " Time: 4.272110939025879 \n",
            "\n",
            " Loss: 0.057638127356767654 \n",
            " Epoch: 26 \n",
            " Time: 3.870828151702881 \n",
            "\n",
            " Loss: 0.05543455481529236 \n",
            " Epoch: 27 \n",
            " Time: 4.3362040519714355 \n",
            "\n",
            " Loss: 0.06183934211730957 \n",
            " Epoch: 28 \n",
            " Time: 4.342348575592041 \n",
            "\n",
            " Loss: 0.06950397044420242 \n",
            " Epoch: 29 \n",
            " Time: 3.790632724761963 \n",
            "\n",
            " Loss: 0.05834464356303215 \n",
            " Epoch: 30 \n",
            " Time: 4.309391498565674 \n",
            "\n",
            " Loss: 0.05682048574090004 \n",
            " Epoch: 31 \n",
            " Time: 4.353807687759399 \n",
            "\n",
            " Loss: 0.057626351714134216 \n",
            " Epoch: 32 \n",
            " Time: 3.8533215522766113 \n",
            "\n",
            " Loss: 0.057365722954273224 \n",
            " Epoch: 33 \n",
            " Time: 4.3447771072387695 \n",
            "\n",
            " Loss: 0.05451442673802376 \n",
            " Epoch: 34 \n",
            " Time: 4.3082115650177 \n",
            "\n",
            " Loss: 0.05703624710440636 \n",
            " Epoch: 35 \n",
            " Time: 3.8795578479766846 \n",
            "\n",
            " Loss: 0.050958920270204544 \n",
            " Epoch: 36 \n",
            " Time: 4.345814943313599 \n",
            "\n",
            " Loss: 0.05606577917933464 \n",
            " Epoch: 37 \n",
            " Time: 4.282556533813477 \n",
            "\n",
            " Loss: 0.05652172118425369 \n",
            " Epoch: 38 \n",
            " Time: 3.840806245803833 \n",
            "\n",
            " Loss: 0.05358635261654854 \n",
            " Epoch: 39 \n",
            " Time: 4.335792779922485 \n",
            "\n",
            " Loss: 0.05688847601413727 \n",
            " Epoch: 40 \n",
            " Time: 4.266405344009399 \n",
            "\n",
            " Loss: 0.05484456941485405 \n",
            " Epoch: 41 \n",
            " Time: 3.7704594135284424 \n",
            "\n",
            " Loss: 0.08122767508029938 \n",
            " Epoch: 42 \n",
            " Time: 4.282357215881348 \n",
            "\n",
            " Loss: 0.05367210507392883 \n",
            " Epoch: 43 \n",
            " Time: 4.31139349937439 \n",
            "\n",
            " Loss: 0.05809972062706947 \n",
            " Epoch: 44 \n",
            " Time: 3.8263041973114014 \n",
            "\n",
            " Loss: 0.05937046557664871 \n",
            " Epoch: 45 \n",
            " Time: 4.207170009613037 \n",
            "\n",
            " Loss: 0.05837051570415497 \n",
            " Epoch: 46 \n",
            " Time: 4.2997846603393555 \n",
            "\n",
            " Loss: 0.06125036999583244 \n",
            " Epoch: 47 \n",
            " Time: 3.7709221839904785 \n",
            "\n",
            " Loss: 0.04988598823547363 \n",
            " Epoch: 48 \n",
            " Time: 4.158764123916626 \n",
            "\n",
            " Loss: 0.0505429245531559 \n",
            " Epoch: 49 \n",
            " Time: 4.3661041259765625 \n",
            "\n",
            " Loss: 0.04752613604068756 \n",
            " Epoch: 50 \n",
            " Time: 3.6897494792938232 \n",
            "\n",
            " Loss: 0.04809184372425079 \n",
            " Epoch: 51 \n",
            " Time: 3.864471435546875 \n",
            "\n",
            " Loss: 0.05853896588087082 \n",
            " Epoch: 52 \n",
            " Time: 4.710219860076904 \n",
            "\n",
            " Loss: 0.05588236078619957 \n",
            " Epoch: 53 \n",
            " Time: 3.7713661193847656 \n",
            "\n",
            " Loss: 0.04458606243133545 \n",
            " Epoch: 54 \n",
            " Time: 3.8079888820648193 \n",
            "\n",
            " Loss: 0.05576581507921219 \n",
            " Epoch: 55 \n",
            " Time: 4.790109634399414 \n",
            "\n",
            " Loss: 0.04999168589711189 \n",
            " Epoch: 56 \n",
            " Time: 3.8188605308532715 \n",
            "\n",
            " Loss: 0.05998624861240387 \n",
            " Epoch: 57 \n",
            " Time: 3.823514699935913 \n",
            "\n",
            " Loss: 0.0515049509704113 \n",
            " Epoch: 58 \n",
            " Time: 4.790449142456055 \n",
            "\n",
            " Loss: 0.051221951842308044 \n",
            " Epoch: 59 \n",
            " Time: 3.748666524887085 \n",
            "\n",
            " Loss: 0.05232115462422371 \n",
            " Epoch: 60 \n",
            " Time: 3.849761962890625 \n",
            "\n",
            " Loss: 0.04946895316243172 \n",
            " Epoch: 61 \n",
            " Time: 4.771920919418335 \n",
            "\n",
            " Loss: 0.0529966801404953 \n",
            " Epoch: 62 \n",
            " Time: 3.8133885860443115 \n",
            "\n",
            " Loss: 0.056800082325935364 \n",
            " Epoch: 63 \n",
            " Time: 3.906832695007324 \n",
            "\n",
            " Loss: 0.04465813934803009 \n",
            " Epoch: 64 \n",
            " Time: 4.782098770141602 \n",
            "\n",
            " Loss: 0.045270130038261414 \n",
            " Epoch: 65 \n",
            " Time: 3.769376754760742 \n",
            "\n",
            " Loss: 0.054189812391996384 \n",
            " Epoch: 66 \n",
            " Time: 3.787750720977783 \n",
            "\n",
            " Loss: 0.04769529774785042 \n",
            " Epoch: 67 \n",
            " Time: 4.739227294921875 \n",
            "\n",
            " Loss: 0.04238996282219887 \n",
            " Epoch: 68 \n",
            " Time: 3.805596113204956 \n",
            "\n",
            " Loss: 0.04814906790852547 \n",
            " Epoch: 69 \n",
            " Time: 3.881166458129883 \n",
            "\n",
            " Loss: 0.05420231446623802 \n",
            " Epoch: 70 \n",
            " Time: 4.863345146179199 \n",
            "\n",
            " Loss: 0.0592031329870224 \n",
            " Epoch: 71 \n",
            " Time: 3.802802562713623 \n",
            "\n",
            " Loss: 0.06278257817029953 \n",
            " Epoch: 72 \n",
            " Time: 3.828540802001953 \n",
            "\n",
            " Loss: 0.06040169298648834 \n",
            " Epoch: 73 \n",
            " Time: 4.791761875152588 \n",
            "\n",
            " Loss: 0.05188112333416939 \n",
            " Epoch: 74 \n",
            " Time: 3.7710330486297607 \n",
            "\n",
            " Loss: 0.04990224540233612 \n",
            " Epoch: 75 \n",
            " Time: 3.785517454147339 \n",
            "\n",
            " Loss: 0.057023320347070694 \n",
            " Epoch: 76 \n",
            " Time: 4.739245653152466 \n",
            "\n",
            " Loss: 0.05004794895648956 \n",
            " Epoch: 77 \n",
            " Time: 3.7841944694519043 \n",
            "\n",
            " Loss: 0.04475074261426926 \n",
            " Epoch: 78 \n",
            " Time: 3.8536875247955322 \n",
            "\n",
            " Loss: 0.05170156806707382 \n",
            " Epoch: 79 \n",
            " Time: 4.767796754837036 \n",
            "\n",
            " Loss: 0.04840974882245064 \n",
            " Epoch: 80 \n",
            " Time: 3.7918026447296143 \n",
            "\n",
            " Loss: 0.046912334859371185 \n",
            " Epoch: 81 \n",
            " Time: 3.847726583480835 \n",
            "\n",
            " Loss: 0.04501071944832802 \n",
            " Epoch: 82 \n",
            " Time: 4.841584920883179 \n",
            "\n",
            " Loss: 0.04641281068325043 \n",
            " Epoch: 83 \n",
            " Time: 3.8615365028381348 \n",
            "\n",
            " Loss: 0.04945041239261627 \n",
            " Epoch: 84 \n",
            " Time: 3.875030994415283 \n",
            "\n",
            " Loss: 0.053782328963279724 \n",
            " Epoch: 85 \n",
            " Time: 4.8645429611206055 \n",
            "\n",
            " Loss: 0.048454705625772476 \n",
            " Epoch: 86 \n",
            " Time: 3.7933578491210938 \n",
            "\n",
            " Loss: 0.05263875424861908 \n",
            " Epoch: 87 \n",
            " Time: 3.7839736938476562 \n",
            "\n",
            " Loss: 0.04639091342687607 \n",
            " Epoch: 88 \n",
            " Time: 4.765676021575928 \n",
            "\n",
            " Loss: 0.044775672256946564 \n",
            " Epoch: 89 \n",
            " Time: 3.800405502319336 \n",
            "\n",
            " Loss: 0.04739949479699135 \n",
            " Epoch: 90 \n",
            " Time: 3.8082833290100098 \n",
            "\n",
            " Loss: 0.04526563361287117 \n",
            " Epoch: 91 \n",
            " Time: 4.747692346572876 \n",
            "\n",
            " Loss: 0.05037842318415642 \n",
            " Epoch: 92 \n",
            " Time: 3.805635929107666 \n",
            "\n",
            " Loss: 0.052029624581336975 \n",
            " Epoch: 93 \n",
            " Time: 3.923069953918457 \n",
            "\n",
            " Loss: 0.055726416409015656 \n",
            " Epoch: 94 \n",
            " Time: 4.842767953872681 \n",
            "\n",
            " Loss: 0.06814292818307877 \n",
            " Epoch: 95 \n",
            " Time: 3.800743579864502 \n",
            "\n",
            " Loss: 0.04592014476656914 \n",
            " Epoch: 96 \n",
            " Time: 3.806522846221924 \n",
            "\n",
            " Loss: 0.051221173256635666 \n",
            " Epoch: 97 \n",
            " Time: 4.7906553745269775 \n",
            "\n",
            " Loss: 0.05430305004119873 \n",
            " Epoch: 98 \n",
            " Time: 3.841480016708374 \n",
            "\n",
            " Loss: 0.04675285145640373 \n",
            " Epoch: 99 \n",
            " Time: 3.8728790283203125 \n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction = model(x_test_lr[:1])\n",
        "prediction = prediction.permute(0, 2, 3, 1)\n",
        "\n",
        "plt.imshow(prediction[0].detach().numpy())\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "rXOZBdpTZC0W",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 467
        },
        "outputId": "d9c0c0b1-d554-42c3-9336-b4b96c74fb69"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "true_im = x_test_hr[:1]\n",
        "true_im = true_im.permute(0, 2, 3, 1)\n",
        "\n",
        "plt.imshow(true_im[0].detach().numpy())\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "en6Xc3aKEVrz",
        "outputId": "4c78f534-cfed-4759-88dd-3a7c8e5220bd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "input_im = x_test_lr[:1]\n",
        "input_im = input_im.permute(0, 2, 3, 1)\n",
        "\n",
        "plt.imshow(input_im[0].detach().numpy())\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "OYO2SZ_7FBg3",
        "outputId": "b3f6598d-96d0-4024-ac6b-55f6134654e7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}