Dataloaders_WandB.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/felixjay3/planets/blob/main/Dataloaders_WandB.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gK4aLhBA5vgN"
      },
      "source": [
        "# DataLoaders - Reading, Writing, Saving\n",
        "## Reading & Saving"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "\n",
        "# npy\n",
        "\n",
        "tmp_array = np.ones((3,3))\n",
        "np.save(\"tmp_array.npy\", tmp_array)\n",
        "read_array = np.load(\"tmp_array.npy\")\n",
        "\n",
        "tmp_array_pkl = np.array([[0,1],[2,3,4],[5,6,7,8]], dtype=object)\n",
        "np.save(\"tmp_array_pkl.npy\", tmp_array_pkl, allow_pickle=True)\n",
        "read_array_pkl = np.load(\"tmp_array_pkl.npy\", allow_pickle=True)\n",
        "\n",
        "print(read_array)\n",
        "print(read_array_pkl)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9AEFoKRnHqV9",
        "outputId": "8d53c631-a704-497e-97e1-7cfb2ba2c613"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[1. 1. 1.]\n",
            " [1. 1. 1.]\n",
            " [1. 1. 1.]]\n",
            "[list([0, 1]) list([2, 3, 4]) list([5, 6, 7, 8])]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# npz\n",
        "np.savez('tmp.npz', tmp_array=tmp_array, tmp_array_pkl=tmp_array_pkl)\n",
        "npz_data = np.load('tmp.npz', allow_pickle=True, encoding='bytes')\n",
        "read_array = npz_data['tmp_array']\n",
        "read_array_pkl = npz_data['tmp_array_pkl']\n",
        "\n",
        "print(read_array)\n",
        "print(read_array_pkl)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_n3kip7VK8qP",
        "outputId": "530f5515-cce4-4cbd-f4b6-2382eed639a3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[1. 1. 1.]\n",
            " [1. 1. 1.]\n",
            " [1. 1. 1.]]\n",
            "[list([0, 1]) list([2, 3, 4]) list([5, 6, 7, 8])]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# csv\n",
        "import pandas as pd\n",
        "\n",
        "output = pd.DataFrame()\n",
        "output['id'] = np.array(range(10))\n",
        "output['label'] = np.array(range(10,20))\n",
        "print(output.head())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qIILXNgxK8zl",
        "outputId": "b42981f9-28d5-4d5d-8fec-5c38f44ef067"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "   id  label\n",
            "0   0     10\n",
            "1   1     11\n",
            "2   2     12\n",
            "3   3     13\n",
            "4   4     14\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "output.to_csv(\"submission.csv\", index = False)\n",
        "\n",
        "output_read = pd.read_csv(\"submission.csv\")\n",
        "print(output_read.head())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K32NDtxtWNZx",
        "outputId": "529cb109-9cac-46a7-d67f-65139ca84a2b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "   id  label\n",
            "0   0     10\n",
            "1   1     11\n",
            "2   2     12\n",
            "3   3     13\n",
            "4   4     14\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Custom Dataset & Dataloader"
      ],
      "metadata": {
        "id": "WhM8EFRyHstP"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Q4xLCpiUscQl",
        "outputId": "9dd88390-418b-41b2-dcc5-c956a0757a49"
      },
      "source": [
        "xs = list(range(10))\n",
        "ys = list(range(10,20))\n",
        "print('xs values: ', xs)\n",
        "print('ys values: ', ys)"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "xs values:  [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]\n",
            "ys values:  [10, 11, 12, 13, 14, 15, 16, 17, 18, 19]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bOvoSj9yscDL",
        "outputId": "12d9cac7-f2eb-43e9-c31e-cf79e216249f"
      },
      "source": [
        "data = list(zip(xs,ys))\n",
        "data[0] # returns the tuple (x[0], y[0])"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FUu9t0uuoJp6",
        "outputId": "28057736-fa57-4ef8-f9b9-d1ae3f2911b5"
      },
      "source": [
        "len(data)"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "10"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ssoP34npsb1A"
      },
      "source": [
        "from torch.utils.data import Dataset, DataLoader\n",
        "\n",
        "class MyDataset(Dataset):\n",
        "    def __init__(self, xs, ys):\n",
        "        self.xs = xs\n",
        "        self.ys = ys\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.xs)  #len(self.ys)\n",
        "\n",
        "    def __getitem__(self, i):\n",
        "        return self.xs[i], self.ys[i]"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qXVsfXfZswm0",
        "outputId": "dd658f76-867b-426f-e2b9-da1b319c41de"
      },
      "source": [
        "dataset = MyDataset(xs, ys)\n",
        "dataset[0] # returns the tuple (x[2], y[2])"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eJWGIftas42A",
        "outputId": "e4cd07ce-32f9-42df-a668-30b62266bdea"
      },
      "source": [
        "from torch.utils.data import DataLoader\n",
        "\n",
        "for x, y in DataLoader(dataset):\n",
        "    print(x,y)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([0]) tensor([10])\n",
            "tensor([1]) tensor([11])\n",
            "tensor([2]) tensor([12])\n",
            "tensor([3]) tensor([13])\n",
            "tensor([4]) tensor([14])\n",
            "tensor([5]) tensor([15])\n",
            "tensor([6]) tensor([16])\n",
            "tensor([7]) tensor([17])\n",
            "tensor([8]) tensor([18])\n",
            "tensor([9]) tensor([19])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3bU-x17Bs7aP",
        "outputId": "b9dea694-e23c-45cf-b514-5422b7f0327b"
      },
      "source": [
        "for x, y in DataLoader(dataset, batch_size=2):\n",
        "    print(x)\n",
        "    print(y)\n",
        "    print(x.shape)\n",
        "    print(y.shape)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([0, 1])\n",
            "tensor([10, 11])\n",
            "torch.Size([2])\n",
            "torch.Size([2])\n",
            "tensor([2, 3])\n",
            "tensor([12, 13])\n",
            "torch.Size([2])\n",
            "torch.Size([2])\n",
            "tensor([4, 5])\n",
            "tensor([14, 15])\n",
            "torch.Size([2])\n",
            "torch.Size([2])\n",
            "tensor([6, 7])\n",
            "tensor([16, 17])\n",
            "torch.Size([2])\n",
            "torch.Size([2])\n",
            "tensor([8, 9])\n",
            "tensor([18, 19])\n",
            "torch.Size([2])\n",
            "torch.Size([2])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EyIyOs_dtEW-",
        "outputId": "c0b1834b-d0c8-44b7-8f7f-833f49b7f204"
      },
      "source": [
        "# enumerate\n",
        "for index, (x, y) in enumerate(DataLoader(dataset, batch_size=2, shuffle=True)):\n",
        "    print(\"index: \", index, \"x,y:\", x,y)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "index:  0 x,y: tensor([5, 2]) tensor([15, 12])\n",
            "index:  1 x,y: tensor([7, 3]) tensor([17, 13])\n",
            "index:  2 x,y: tensor([0, 9]) tensor([10, 19])\n",
            "index:  3 x,y: tensor([4, 6]) tensor([14, 16])\n",
            "index:  4 x,y: tensor([8, 1]) tensor([18, 11])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Training\n",
        "train_loader_args = dict(shuffle=True, batch_size=2, num_workers=2, pin_memory=True)\n",
        "train_loader = DataLoader(dataset, **train_loader_args)"
      ],
      "metadata": {
        "id": "pfIdnr6YYa3A"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# enumerate\n",
        "for index, (x, y) in enumerate(train_loader):\n",
        "    print(\"index: \", index, \"x,y:\", x,y)"
      ],
      "metadata": {
        "id": "QlTFJs4g17NQ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a57287da-292e-41a6-f47c-2046e79c3890"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "index:  0 x,y: tensor([7, 6]) tensor([17, 16])\n",
            "index:  1 x,y: tensor([4, 2]) tensor([14, 12])\n",
            "index:  2 x,y: tensor([5, 3]) tensor([15, 13])\n",
            "index:  3 x,y: tensor([9, 0]) tensor([19, 10])\n",
            "index:  4 x,y: tensor([1, 8]) tensor([11, 18])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Torch Dataset & Dataloader"
      ],
      "metadata": {
        "id": "TyJ8oG-0106b"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gibYD_zK42gc"
      },
      "source": [
        "import torch\n",
        "from torch.utils.data import Dataset\n",
        "from torchvision import datasets\n",
        "from torchvision.transforms import ToTensor\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# FashionMNIST dataset from TorchVision\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",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 675
        },
        "id": "Z8WnGm675PLi",
        "outputId": "4bfb30f9-40c8-4223-c7d6-5d5281fd008c"
      },
      "source": [
        "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",
        "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()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x800 with 9 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BENUTMfE8qUB"
      },
      "source": [
        "from torch.utils.data import DataLoader\n",
        "\n",
        "train_dataloader = DataLoader(training_data, batch_size=64, shuffle=True)\n",
        "test_dataloader = DataLoader(test_data, batch_size=64, shuffle=False)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 517
        },
        "id": "aqe-LKwp8e_S",
        "outputId": "4dec4ca5-b0c3-4832-a629-7cf9f59206a2"
      },
      "source": [
        "# Display image and label.\n",
        "train_features, train_labels = next(iter(train_dataloader))\n",
        "print(f\"Feature batch shape: {train_features.size()}\")\n",
        "print(f\"Labels batch shape: {train_labels.size()}\")\n",
        "print(train_features[0].shape)\n",
        "img = train_features[0].squeeze()\n",
        "print(img.shape)\n",
        "label = train_labels[0]\n",
        "plt.imshow(img, cmap=\"gray\")\n",
        "plt.show()\n",
        "print(f\"Label: {label}\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Feature batch shape: torch.Size([64, 1, 28, 28])\n",
            "Labels batch shape: torch.Size([64])\n",
            "torch.Size([1, 28, 28])\n",
            "torch.Size([28, 28])\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Label: 0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Custom Image Dataset"
      ],
      "metadata": {
        "id": "byUFrzff3Hjx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Save MNIST images in zip and labels in annotations file\n",
        "\n",
        "train_images_custom = []\n",
        "train_labels_custom = []\n",
        "for i in range(10):\n",
        "  train_images_custom.append(training_data[i][0].numpy())\n",
        "  train_labels_custom.append(training_data[i][1])\n",
        "\n",
        "images_npy = np.array(train_images_custom)\n",
        "np.save(\"images_npy.npy\", images_npy)\n",
        "\n",
        "labels_npy = np.array(train_labels_custom)\n",
        "np.save(\"labels_npy.npy\", labels_npy)\n"
      ],
      "metadata": {
        "id": "oJ6zb_PKPIWT"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EepabNTy5WK3"
      },
      "source": [
        "class CustomImageDataset(Dataset):\n",
        "    def __init__(self, images_npy, labels_npy):\n",
        "        self.imgs = np.load(images_npy)\n",
        "        self.img_labels = np.load(labels_npy)\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.img_labels)\n",
        "\n",
        "    def __getitem__(self, idx):\n",
        "        image = self.imgs[idx]\n",
        "        label = self.img_labels[idx]\n",
        "        return image, label\n",
        "\n",
        "    def collate_fn(self,batch):\n",
        "      xs, ys = zip(*batch)\n",
        "\n",
        "      #For demo - padding border\n",
        "      xs = [np.pad(x_i, ((0,0),(20, 20), (20, 20)), constant_values=0) for x_i in list(xs) ]\n",
        "      return tuple(xs), ys\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_data_custom = CustomImageDataset(\"images_npy.npy\", \"labels_npy.npy\")\n",
        "train_args = dict(batch_size=2, shuffle=True, collate_fn= train_data_custom.collate_fn)\n",
        "train_dataloader_custom = DataLoader(train_data_custom, **train_args)\n",
        "\n",
        "# Display image and label.\n",
        "train_features, train_labels = next(iter(train_dataloader_custom))\n",
        "img = train_features[0].squeeze()\n",
        "label = train_labels[0]\n",
        "plt.imshow(img, cmap=\"gray\")\n",
        "plt.show()\n",
        "print(f\"Label: {label}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        },
        "id": "TWxaBNizTHLv",
        "outputId": "9b876fa2-a9cc-47af-daf9-adf893127604"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Label: tensor([18, 10])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\"\"\"Some references:\n",
        "\n",
        "    1. Numpy save and load - https://numpy.org/doc/stable/reference/generated/numpy.save.html\n",
        "                             https://numpy.org/doc/stable/reference/generated/numpy.load.html\n",
        "                             https://numpy.org/doc/stable/reference/generated/numpy.savez.html#numpy.savez\n",
        "\n",
        "    2. Pandas read and save csv - https://pandas.pydata.org/docs/reference/api/pandas.read_csv.html\n",
        "                                  https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.to_csv.html\n",
        "\n",
        "    2. Pytorch Datasets and Datloader - https://pytorch.org/tutorials/beginner/basics/data_tutorial.html\n",
        "                                        https://pytorch.org/docs/stable/data.html\n",
        "\n",
        "\"\"\""
      ],
      "metadata": {
        "id": "IjtRHoSsQ2lf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Monitoring Training using Weights and Biases (WandB)"
      ],
      "metadata": {
        "id": "RFkaFoYuU2TZ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## What is WandB?\n",
        "Weights & Biases (WandB) is a python package that allows us to monitor our training in real-time. It can be easily integrated with popular deep learning frameworks like Pytorch, Tensorflow, or Keras. Additionally, it allows us to organize our Runs into Projects where we can easily compare them and identify the best performing model. In this guide, we will learn how to use WandB for logging."
      ],
      "metadata": {
        "id": "Bmw8sRXOU-wL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "%pip install wandb -q\n",
        "import wandb\n",
        "wandb.login()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 191
        },
        "id": "GLftZ8TRU1fH",
        "outputId": "62702f7a-91d9-494b-e9e0-ee43c37617e7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.2/2.2 MB\u001b[0m \u001b[31m9.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m195.4/195.4 kB\u001b[0m \u001b[31m21.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m258.5/258.5 kB\u001b[0m \u001b[31m24.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m62.7/62.7 kB\u001b[0m \u001b[31m7.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ],
            "application/javascript": [
              "\n",
              "        window._wandbApiKey = new Promise((resolve, reject) => {\n",
              "            function loadScript(url) {\n",
              "            return new Promise(function(resolve, reject) {\n",
              "                let newScript = document.createElement(\"script\");\n",
              "                newScript.onerror = reject;\n",
              "                newScript.onload = resolve;\n",
              "                document.body.appendChild(newScript);\n",
              "                newScript.src = url;\n",
              "            });\n",
              "            }\n",
              "            loadScript(\"https://cdn.jsdelivr.net/npm/postmate/build/postmate.min.js\").then(() => {\n",
              "            const iframe = document.createElement('iframe')\n",
              "            iframe.style.cssText = \"width:0;height:0;border:none\"\n",
              "            document.body.appendChild(iframe)\n",
              "            const handshake = new Postmate({\n",
              "                container: iframe,\n",
              "                url: 'https://wandb.ai/authorize'\n",
              "            });\n",
              "            const timeout = setTimeout(() => reject(\"Couldn't auto authenticate\"), 5000)\n",
              "            handshake.then(function(child) {\n",
              "                child.on('authorize', data => {\n",
              "                    clearTimeout(timeout)\n",
              "                    resolve(data)\n",
              "                });\n",
              "            });\n",
              "            })\n",
              "        });\n",
              "    "
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[34m\u001b[1mwandb\u001b[0m: Logging into wandb.ai. (Learn how to deploy a W&B server locally: https://wandb.me/wandb-server)\n",
            "\u001b[34m\u001b[1mwandb\u001b[0m: You can find your API key in your browser here: https://wandb.ai/authorize\n",
            "wandb: Paste an API key from your profile and hit enter, or press ctrl+c to quit:"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            " ··········\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[34m\u001b[1mwandb\u001b[0m: Appending key for api.wandb.ai to your netrc file: /root/.netrc\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Initialize WandB\n",
        "wandb.init(project='fmnist_project', entity='maomer')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "ZsUjoGcfj_Q4",
        "outputId": "7113e0db-3532-4612-f258-657a53218d8a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[34m\u001b[1mwandb\u001b[0m: Currently logged in as: \u001b[33mmaomer\u001b[0m. Use \u001b[1m`wandb login --relogin`\u001b[0m to force relogin\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Tracking run with wandb version 0.16.3"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Run data is saved locally in <code>/content/wandb/run-20240229_155240-x9n44zsm</code>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Syncing run <strong><a href='https://wandb.ai/maomer/fmnist_project/runs/x9n44zsm' target=\"_blank\">noble-rain-1</a></strong> to <a href='https://wandb.ai/maomer/fmnist_project' target=\"_blank\">Weights & Biases</a> (<a href='https://wandb.me/run' target=\"_blank\">docs</a>)<br/>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              " View project at <a href='https://wandb.ai/maomer/fmnist_project' target=\"_blank\">https://wandb.ai/maomer/fmnist_project</a>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              " View run at <a href='https://wandb.ai/maomer/fmnist_project/runs/x9n44zsm' target=\"_blank\">https://wandb.ai/maomer/fmnist_project/runs/x9n44zsm</a>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<button onClick=\"this.nextSibling.style.display='block';this.style.display='none';\">Display W&B run</button><iframe src='https://wandb.ai/maomer/fmnist_project/runs/x9n44zsm?jupyter=true' style='border:none;width:100%;height:420px;display:none;'></iframe>"
            ],
            "text/plain": [
              "<wandb.sdk.wandb_run.Run at 0x7bc841205c60>"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms\n",
        "\n",
        "\n",
        "transform = transforms.Compose(\n",
        "    [transforms.ToTensor(),\n",
        "    transforms.Normalize((0.5,), (0.5,))])\n",
        "\n",
        "# Create datasets for training & validation, download if necessary\n",
        "training_set = torchvision.datasets.FashionMNIST('./data', train=True, transform=transform, download=True)\n",
        "validation_set = torchvision.datasets.FashionMNIST('./data', train=False, transform=transform, download=True)\n",
        "\n",
        "# Create data loaders for our datasets; shuffle for training, not for validation\n",
        "training_loader = torch.utils.data.DataLoader(training_set, batch_size=60, shuffle=True)\n",
        "validation_loader = torch.utils.data.DataLoader(validation_set, batch_size=10, shuffle=False)\n",
        "\n",
        "# Class labels\n",
        "classes = ('T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',\n",
        "        'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle Boot')\n",
        "\n",
        "# Report split sizes\n",
        "print('Training set has {} instances'.format(len(training_set)))\n",
        "print('Validation set has {} instances'.format(len(validation_set)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I6xx-m3jUzqK",
        "outputId": "eb3f41cb-53db-47fd-b730-6bb47f316644"
      },
      "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, 15674236.41it/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, 270686.35it/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, 4957391.77it/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, 9588044.85it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz to ./data/FashionMNIST/raw\n",
            "\n",
            "Training set has 60000 instances\n",
            "Validation set has 10000 instances\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "\n",
        "# PyTorch models inherit from torch.nn.Module\n",
        "class GarmentClassifier(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(GarmentClassifier, self).__init__()\n",
        "        self.conv1 = nn.Conv2d(1, 6, 5)\n",
        "        self.pool = nn.MaxPool2d(2, 2)\n",
        "        self.conv2 = nn.Conv2d(6, 16, 5)\n",
        "        self.fc1 = nn.Linear(16 * 4 * 4, 120)\n",
        "        self.fc2 = nn.Linear(120, 84)\n",
        "        self.fc3 = nn.Linear(84, 10)\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.pool(F.relu(self.conv1(x)))\n",
        "        x = self.pool(F.relu(self.conv2(x)))\n",
        "        x = x.view(-1, 16 * 4 * 4)\n",
        "        x = F.relu(self.fc1(x))\n",
        "        x = F.relu(self.fc2(x))\n",
        "        x = self.fc3(x)\n",
        "        return x\n",
        "\n",
        "\n",
        "model = GarmentClassifier()"
      ],
      "metadata": {
        "id": "2ogMvxfnbv4I"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "loss_fn = torch.nn.CrossEntropyLoss()\n",
        "\n",
        "# NB: Loss functions expect data in batches, so we're creating batches of 4\n",
        "# Represents the model's confidence in each of the 10 classes for a given input\n",
        "dummy_outputs = torch.rand(4, 10)\n",
        "# Represents the correct class among the 10 being tested\n",
        "dummy_labels = torch.tensor([1, 5, 3, 7])\n",
        "\n",
        "print(dummy_outputs)\n",
        "print(dummy_labels)\n",
        "\n",
        "loss = loss_fn(dummy_outputs, dummy_labels)\n",
        "print('Total loss for this batch: {}'.format(loss.item()))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XyxKJjOXbvzK",
        "outputId": "fabfc5c4-c12e-4cdc-d527-fdf863cd0bb0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([[0.4010, 0.9824, 0.7652, 0.0082, 0.3153, 0.5429, 0.8819, 0.5957, 0.7540,\n",
            "         0.1834],\n",
            "        [0.5030, 0.6923, 0.9700, 0.5327, 0.6282, 0.8563, 0.8078, 0.3734, 0.9512,\n",
            "         0.4024],\n",
            "        [0.1514, 0.0455, 0.7546, 0.4874, 0.3141, 0.7089, 0.0408, 0.3677, 0.4466,\n",
            "         0.7399],\n",
            "        [0.5227, 0.8678, 0.6673, 0.3845, 0.8935, 0.4776, 0.7758, 0.2704, 0.4787,\n",
            "         0.0435]])\n",
            "tensor([1, 5, 3, 7])\n",
            "Total loss for this batch: 2.2255301475524902\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "optimizer = torch.optim.SGD(model.parameters(), lr=0.001, momentum=0.9)"
      ],
      "metadata": {
        "id": "r4KSapECbvpB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "wandb.watch(model, loss_fn)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JWRTq49reSr0",
        "outputId": "5bcb75a6-0388-4ef1-fd1f-bb80f217c4d2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[]"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def train_one_epoch():\n",
        "    running_loss = 0.\n",
        "    last_loss = 0.\n",
        "    correct = 0\n",
        "    prediction_list = []\n",
        "    total = 0\n",
        "    # Here, we use enumerate(training_loader) instead of\n",
        "    # iter(training_loader) so that we can track the batch\n",
        "    # index and do some intra-epoch reporting\n",
        "\n",
        "    for i, data in enumerate(training_loader):\n",
        "        # Every data instance is an input + label pair\n",
        "        inputs, labels = data\n",
        "\n",
        "        # Zero your gradients for every batch!\n",
        "        optimizer.zero_grad()\n",
        "\n",
        "        # Make predictions for this batch\n",
        "        outputs = model(inputs)\n",
        "\n",
        "        # Compute the loss and its gradients\n",
        "        loss = loss_fn(outputs, labels)\n",
        "        loss.backward()\n",
        "\n",
        "        # Adjust learning weights\n",
        "        optimizer.step()\n",
        "\n",
        "        # Gather data and report\n",
        "        running_loss += loss.item()\n",
        "        predictions = torch.max(outputs, 1)[1]\n",
        "        prediction_list.append(predictions)\n",
        "        correct += (predictions == labels).sum()\n",
        "\n",
        "        total += len(labels)\n",
        "\n",
        "    accuracy = correct * 100 / total\n",
        "\n",
        "    avg_loss = running_loss / (i+1) # loss per batch\n",
        "\n",
        "    return avg_loss, accuracy"
      ],
      "metadata": {
        "id": "19_zd1Oec7Ic"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Initializing in a separate cell so we can easily add more epochs to the same run\n",
        "\n",
        "EPOCHS = 100\n",
        "\n",
        "\n",
        "for epoch in range(EPOCHS):\n",
        "    print('EPOCH {}:'.format(epoch + 1))\n",
        "\n",
        "    # Make sure gradient tracking is on, and do a pass over the data\n",
        "    model.train(True)\n",
        "    avg_loss, avg_acc = train_one_epoch()\n",
        "\n",
        "\n",
        "    running_vloss = 0.0\n",
        "    # Set the model to evaluation mode, disabling dropout and using population\n",
        "    # statistics for batch normalization.\n",
        "    model.eval()\n",
        "\n",
        "    # Disable gradient computation and reduce memory consumption.\n",
        "    with torch.no_grad():\n",
        "        prediction_listv = []\n",
        "        correctv = 0\n",
        "        totalv = 0\n",
        "        for i, vdata in enumerate(validation_loader):\n",
        "            vinputs, vlabels = vdata\n",
        "            voutputs = model(vinputs)\n",
        "            vloss = loss_fn(voutputs, vlabels)\n",
        "            running_vloss += vloss\n",
        "            predictionsv = torch.max(voutputs, 1)[1]\n",
        "            prediction_listv.append(predictionsv)\n",
        "            correctv += (predictionsv == vlabels).sum()\n",
        "\n",
        "            totalv += len(vlabels)\n",
        "\n",
        "    accuracy_valid = correctv * 100 / totalv\n",
        "    avg_vloss = running_vloss / (i + 1)\n",
        "    print('LOSS train {} valid {}'.format(avg_loss, avg_vloss))\n",
        "    print('ACC train {} valid {}'.format(avg_acc, accuracy_valid))\n",
        "\n",
        "\n",
        "    # Log the running loss averaged per batch\n",
        "    # for both training and validation\n",
        "    wandb.log({\n",
        "        \"Epoch\": (epoch+1),\n",
        "        \"Train Loss\": avg_loss,\n",
        "        \"Valid Loss\": avg_vloss,\n",
        "        \"Train Acc\":avg_acc,\n",
        "        \"Valid Acc\": accuracy_valid})\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "449Gnn0WcUPS",
        "outputId": "4ba453f9-91ce-4d61-a0e3-57df77328598"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "EPOCH 1:\n",
            "LOSS train 1.7064150984883308 valid 0.7987259030342102\n",
            "ACC train 36.91999816894531 valid 70.75\n",
            "EPOCH 2:\n",
            "LOSS train 0.6776801473796368 valid 0.6372979283332825\n",
            "ACC train 74.23666381835938 valid 75.75\n",
            "EPOCH 3:\n",
            "LOSS train 0.5720613241195679 valid 0.5628799796104431\n",
            "ACC train 78.22333526611328 valid 78.41000366210938\n",
            "EPOCH 4:\n",
            "LOSS train 0.5163080155998468 valid 0.5175269842147827\n",
            "ACC train 80.78166961669922 valid 80.68000030517578\n",
            "EPOCH 5:\n",
            "LOSS train 0.47426905170083045 valid 0.48243290185928345\n",
            "ACC train 82.45500183105469 valid 82.08999633789062\n",
            "EPOCH 6:\n",
            "LOSS train 0.4443372343480587 valid 0.4569040536880493\n",
            "ACC train 83.83000183105469 valid 83.55999755859375\n",
            "EPOCH 7:\n",
            "LOSS train 0.422935411542654 valid 0.4673013687133789\n",
            "ACC train 84.57833099365234 valid 82.62999725341797\n",
            "EPOCH 8:\n",
            "LOSS train 0.40185358487069606 valid 0.4235183000564575\n",
            "ACC train 85.34166717529297 valid 85.12999725341797\n",
            "EPOCH 9:\n",
            "LOSS train 0.38599198842048643 valid 0.4035215377807617\n",
            "ACC train 85.95333099365234 valid 85.55999755859375\n",
            "EPOCH 10:\n",
            "LOSS train 0.3719299981445074 valid 0.4059111773967743\n",
            "ACC train 86.41666412353516 valid 85.26000213623047\n",
            "EPOCH 11:\n",
            "LOSS train 0.3595242169946432 valid 0.39621537923812866\n",
            "ACC train 86.82166290283203 valid 85.66000366210938\n",
            "EPOCH 12:\n",
            "LOSS train 0.34830681336671115 valid 0.37402263283729553\n",
            "ACC train 87.37833404541016 valid 86.70999908447266\n",
            "EPOCH 13:\n",
            "LOSS train 0.3389585639238358 valid 0.37471869587898254\n",
            "ACC train 87.66666412353516 valid 86.13999938964844\n",
            "EPOCH 14:\n",
            "LOSS train 0.32881306947767736 valid 0.36374956369400024\n",
            "ACC train 87.96833038330078 valid 86.93000030517578\n",
            "EPOCH 15:\n",
            "LOSS train 0.32068193749338386 valid 0.35613003373146057\n",
            "ACC train 88.23666381835938 valid 87.0999984741211\n",
            "EPOCH 16:\n",
            "LOSS train 0.31437407752871516 valid 0.34757113456726074\n",
            "ACC train 88.42666625976562 valid 87.25\n",
            "EPOCH 17:\n",
            "LOSS train 0.30711645831912754 valid 0.35017499327659607\n",
            "ACC train 88.77166748046875 valid 87.48999786376953\n",
            "EPOCH 18:\n",
            "LOSS train 0.30248775351792573 valid 0.3402721583843231\n",
            "ACC train 88.88166809082031 valid 87.5\n",
            "EPOCH 19:\n",
            "LOSS train 0.29544524911791087 valid 0.33757996559143066\n",
            "ACC train 89.09333038330078 valid 87.83999633789062\n",
            "EPOCH 20:\n",
            "LOSS train 0.2921379471719265 valid 0.32696494460105896\n",
            "ACC train 89.26499938964844 valid 88.27999877929688\n",
            "EPOCH 21:\n",
            "LOSS train 0.2848629826381803 valid 0.33352571725845337\n",
            "ACC train 89.5250015258789 valid 87.83000183105469\n",
            "EPOCH 22:\n",
            "LOSS train 0.282752475194633 valid 0.3294554650783539\n",
            "ACC train 89.6066665649414 valid 88.16999816894531\n",
            "EPOCH 23:\n",
            "LOSS train 0.275346771068871 valid 0.32041123509407043\n",
            "ACC train 89.91000366210938 valid 88.30000305175781\n",
            "EPOCH 24:\n",
            "LOSS train 0.27139068476483225 valid 0.33010849356651306\n",
            "ACC train 89.97333526611328 valid 88.0199966430664\n",
            "EPOCH 25:\n",
            "LOSS train 0.2675428428798914 valid 0.31405776739120483\n",
            "ACC train 90.163330078125 valid 88.54000091552734\n",
            "EPOCH 26:\n"
          ]
        },
        {
          "output_type": "error",
          "ename": "KeyboardInterrupt",
          "evalue": "",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-20-6b8644b53907>\u001b[0m in \u001b[0;36m<cell line: 6>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     22\u001b[0m         \u001b[0mcorrectv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     23\u001b[0m         \u001b[0mtotalv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 24\u001b[0;31m         \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvdata\u001b[0m \u001b[0;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalidation_loader\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     25\u001b[0m             \u001b[0mvinputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvlabels\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mvdata\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     26\u001b[0m             \u001b[0mvoutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvinputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py\u001b[0m in \u001b[0;36m__next__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    628\u001b[0m                 \u001b[0;31m# TODO(https://github.com/pytorch/pytorch/issues/76750)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    629\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_reset\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# type: ignore[call-arg]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 630\u001b[0;31m             \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_next_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    631\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_num_yielded\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    632\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_dataset_kind\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0m_DatasetKind\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mIterable\u001b[0m \u001b[0;32mand\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m\\\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/torch/utils/data/dataloader.py\u001b[0m in \u001b[0;36m_next_data\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    672\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_next_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    673\u001b[0m         \u001b[0mindex\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_next_index\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# may raise StopIteration\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 674\u001b[0;31m         \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_dataset_fetcher\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfetch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# may raise StopIteration\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    675\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_pin_memory\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    676\u001b[0m             \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_utils\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpin_memory\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpin_memory\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_pin_memory_device\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/fetch.py\u001b[0m in \u001b[0;36mfetch\u001b[0;34m(self, possibly_batched_index)\u001b[0m\n\u001b[1;32m     49\u001b[0m                 \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__getitems__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpossibly_batched_index\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     50\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 51\u001b[0;31m                 \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0midx\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0midx\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mpossibly_batched_index\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     52\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     53\u001b[0m             \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mpossibly_batched_index\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/torch/utils/data/_utils/fetch.py\u001b[0m in \u001b[0;36m<listcomp>\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m     49\u001b[0m                 \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__getitems__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpossibly_batched_index\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     50\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 51\u001b[0;31m                 \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0midx\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0midx\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mpossibly_batched_index\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     52\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     53\u001b[0m             \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mpossibly_batched_index\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/torchvision/datasets/mnist.py\u001b[0m in \u001b[0;36m__getitem__\u001b[0;34m(self, index)\u001b[0m\n\u001b[1;32m    143\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    144\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransform\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 145\u001b[0;31m             \u001b[0mimg\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    146\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    147\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtarget_transform\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/torchvision/transforms/transforms.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, img)\u001b[0m\n\u001b[1;32m     93\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mimg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     94\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mt\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransforms\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 95\u001b[0;31m             \u001b[0mimg\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mt\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     96\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mimg\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     97\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/torchvision/transforms/transforms.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, pic)\u001b[0m\n\u001b[1;32m    135\u001b[0m             \u001b[0mTensor\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mConverted\u001b[0m \u001b[0mimage\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    136\u001b[0m         \"\"\"\n\u001b[0;32m--> 137\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mF\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto_tensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpic\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    138\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    139\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__repr__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mstr\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/torchvision/transforms/functional.py\u001b[0m in \u001b[0;36mto_tensor\u001b[0;34m(pic)\u001b[0m\n\u001b[1;32m    164\u001b[0m     \u001b[0;31m# handle PIL Image\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    165\u001b[0m     \u001b[0mmode_to_nptype\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m{\u001b[0m\u001b[0;34m\"I\"\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mint32\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"I;16\"\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mint16\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"F\"\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat32\u001b[0m\u001b[0;34m}\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 166\u001b[0;31m     \u001b[0mimg\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfrom_numpy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpic\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmode_to_nptype\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpic\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmode\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0muint8\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    167\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    168\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mpic\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmode\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m\"1\"\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/PIL/Image.py\u001b[0m in \u001b[0;36m__array_interface__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    699\u001b[0m                 \u001b[0mnew\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"data\"\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtobytes\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"raw\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"L\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    700\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 701\u001b[0;31m                 \u001b[0mnew\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"data\"\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtobytes\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    702\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    703\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mMemoryError\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mRecursionError\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/PIL/Image.py\u001b[0m in \u001b[0;36mtobytes\u001b[0;34m(self, encoder_name, *args)\u001b[0m\n\u001b[1;32m    763\u001b[0m         \u001b[0;31m# unpack data\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    764\u001b[0m         \u001b[0me\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_getencoder\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmode\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mencoder_name\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 765\u001b[0;31m         \u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msetimage\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mim\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    766\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    767\u001b[0m         \u001b[0mbufsize\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m65536\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0;36m4\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# see RawEncode.c\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "OUvBo03vqUD4"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}