Lhy_Machine_Learning/范例/Pytorch/Pytorch_Tutorial.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Pytorch Tutorial",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ga642381/ML2021-Spring/blob/main/Pytorch/Pytorch_Tutorial.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tHILOGjOQbsQ"
      },
      "source": [
        "# **Pytorch Tutorial**\r\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C1zA7GupxdJv"
      },
      "source": [
        "import torch"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6Eqj90EkWbWx"
      },
      "source": [
        "**1. Pytorch Documentation Explanation with torch.max**\r\n",
        "\r\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JCXOg-iSQuk7"
      },
      "source": [
        "x = torch.randn(4,5)\r\n",
        "y = torch.randn(4,5)\r\n",
        "z = torch.randn(4,5)\r\n",
        "print(x)\r\n",
        "print(y)\r\n",
        "print(z)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EEqa9GFoWF78"
      },
      "source": [
        "# 1. max of entire tensor (torch.max(input) → Tensor)\r\n",
        "m = torch.max(x)\r\n",
        "print(m)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wffThGDyWKxJ"
      },
      "source": [
        "# 2. max along a dimension (torch.max(input, dim, keepdim=False, *, out=None) → (Tensor, LongTensor))\r\n",
        "m, idx = torch.max(x,0)\r\n",
        "print(m)\r\n",
        "print(idx)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oKDQW3tIXKg-"
      },
      "source": [
        "# 2-2\r\n",
        "m, idx = torch.max(input=x,dim=0)\r\n",
        "print(m)\r\n",
        "print(idx)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6QZ6WRLyX3De"
      },
      "source": [
        "# 2-3\r\n",
        "m, idx = torch.max(x,0,False)\r\n",
        "print(m)\r\n",
        "print(idx)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nqGuctkKbUEn"
      },
      "source": [
        "# 2-4\r\n",
        "m, idx = torch.max(x,dim=0,keepdim=True)\r\n",
        "print(m)\r\n",
        "print(idx)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9OMzxuMlZPIu"
      },
      "source": [
        "# 2-5\r\n",
        "p = (m,idx)\r\n",
        "torch.max(x,0,False,out=p)\r\n",
        "print(p[0])\r\n",
        "print(p[1])\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uhd4TqGTbD2c"
      },
      "source": [
        "# 2-6\r\n",
        "p = (m,idx)\r\n",
        "torch.max(x,0,False,p)\r\n",
        "print(p[0])\r\n",
        "print(p[1])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wbxjUSOXxN0n"
      },
      "source": [
        "# 2-7\r\n",
        "m, idx = torch.max(x,True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iMwhGLlGWYaR"
      },
      "source": [
        "# 3. max(choose max) operators on two tensors (torch.max(input, other, *, out=None) → Tensor)\r\n",
        "t = torch.max(x,y)\r\n",
        "print(t)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nFxRKu2Dedwb"
      },
      "source": [
        "**2. Common errors**\r\n",
        "\r\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KMcRyMxGwhul"
      },
      "source": [
        "The following code blocks show some common errors while using the torch library. First, execute the code with error, and then execute the next code block to fix the error. You need to change the runtime to GPU.\r\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eX-kKdi6ynFf"
      },
      "source": [
        "import torch"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-muJ4KKreoP2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        },
        "outputId": "c1d5c3a5-9540-4145-d80c-3cbca18a1deb"
      },
      "source": [
        "# 1. different device error\r\n",
        "model = torch.nn.Linear(5,1).to(\"cuda:0\")\r\n",
        "x = torch.Tensor([1,2,3,4,5]).to(\"cpu\")\r\n",
        "y = model(x)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "RuntimeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-2-12e5b7d55705>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mLinear\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"cuda:0\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m3\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m4\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"cpu\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    726\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--> 727\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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    728\u001b[0m         for hook in itertools.chain(\n\u001b[1;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\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/python3.7/dist-packages/torch/nn/modules/linear.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m     91\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     92\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 93\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mF\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlinear\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbias\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     94\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     95\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mextra_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/python3.7/dist-packages/torch/nn/functional.py\u001b[0m in \u001b[0;36mlinear\u001b[0;34m(input, weight, bias)\u001b[0m\n\u001b[1;32m   1690\u001b[0m         \u001b[0mret\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0maddmm\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbias\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mt\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[1;32m   1691\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-> 1692\u001b[0;31m         \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mweight\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mt\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[0m\u001b[1;32m   1693\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mbias\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[1;32m   1694\u001b[0m             \u001b[0moutput\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0mbias\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mRuntimeError\u001b[0m: Tensor for 'out' is on CPU, Tensor for argument #1 'self' is on CPU, but expected them to be on GPU (while checking arguments for addmm)"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a54PqxJLe9-c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "909d3693-236f-4419-f269-8fb443ef7534"
      },
      "source": [
        "# 1. different device error (fixed)\r\n",
        "x = torch.Tensor([1,2,3,4,5]).to(\"cuda:0\")\r\n",
        "y = model(x)\r\n",
        "print(y.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "torch.Size([1])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n7OHtZwbi7Qw",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 201
        },
        "outputId": "2a7d2dd0-6498-4da0-9591-3554c1739046"
      },
      "source": [
        "# 2. mismatched dimensions error\r\n",
        "x = torch.randn(4,5)\r\n",
        "y= torch.randn(5,4)\r\n",
        "z = x + y"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "RuntimeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-4-7fa8b244df3c>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m4\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m4\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0mz\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mRuntimeError\u001b[0m: The size of tensor a (5) must match the size of tensor b (4) at non-singleton dimension 1"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qVynzvrskFCD",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "926dc01c-be6f-48e1-ad39-a5bcecebc513"
      },
      "source": [
        "# 2. mismatched dimensions error (fixed)\r\n",
        "y= y.transpose(0,1)\r\n",
        "z = x + y\r\n",
        "print(z.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "torch.Size([4, 5])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hgzgb9gJANod",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 398
        },
        "outputId": "21b58850-b3f1-4f2a-db5d-cc45e47ccbea"
      },
      "source": [
        "# 3. cuda out of memory error\n",
        "import torch\n",
        "import torchvision.models as models\n",
        "resnet18 = models.resnet18().to(\"cuda:0\") # Neural Networks for Image Recognition\n",
        "data = torch.randn(2048,3,244,244) # Create fake data (512 images)\n",
        "out = resnet18(data.to(\"cuda:0\")) # Use Data as Input and Feed to Model\n",
        "print(out.shape)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "RuntimeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-8-711923c7f347>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0mresnet18\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodels\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mresnet18\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"cuda:0\"\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Neural Networks for Image Recognition\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m2048\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m3\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m244\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m244\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Create fake data (512 images)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 6\u001b[0;31m \u001b[0mout\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mresnet18\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"cuda:0\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Use Data as Input and Feed to Model\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      7\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mout\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\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/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    726\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--> 727\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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    728\u001b[0m         for hook in itertools.chain(\n\u001b[1;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\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/python3.7/dist-packages/torchvision/models/resnet.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    218\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    219\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\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--> 220\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_forward_impl\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\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    221\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    222\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torchvision/models/resnet.py\u001b[0m in \u001b[0;36m_forward_impl\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    202\u001b[0m         \u001b[0;31m# See note [TorchScript super()]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    203\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconv1\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 204\u001b[0;31m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbn1\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\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    205\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrelu\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    206\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmaxpool\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\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/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    726\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--> 727\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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    728\u001b[0m         for hook in itertools.chain(\n\u001b[1;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\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/python3.7/dist-packages/torch/nn/modules/batchnorm.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m    134\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrunning_mean\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtraining\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrack_running_stats\u001b[0m \u001b[0;32melse\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[1;32m    135\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrunning_var\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtraining\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrack_running_stats\u001b[0m \u001b[0;32melse\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--> 136\u001b[0;31m             self.weight, self.bias, bn_training, exponential_average_factor, self.eps)\n\u001b[0m\u001b[1;32m    137\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    138\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py\u001b[0m in \u001b[0;36mbatch_norm\u001b[0;34m(input, running_mean, running_var, weight, bias, training, momentum, eps)\u001b[0m\n\u001b[1;32m   2056\u001b[0m     return torch.batch_norm(\n\u001b[1;32m   2057\u001b[0m         \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbias\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrunning_mean\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrunning_var\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2058\u001b[0;31m         \u001b[0mtraining\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmomentum\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0meps\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackends\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcudnn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0menabled\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2059\u001b[0m     )\n\u001b[1;32m   2060\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mRuntimeError\u001b[0m: CUDA out of memory. Tried to allocate 7.27 GiB (GPU 0; 14.76 GiB total capacity; 8.74 GiB already allocated; 4.42 GiB free; 9.42 GiB reserved in total by PyTorch)"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VPksKnB_w343",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fbee46ad-e63e-4bfc-8971-452895dd7a15"
      },
      "source": [
        "# 3. cuda out of memory error (fixed)\n",
        "for d in data:\n",
        "  out = resnet18(d.to(\"cuda:0\").unsqueeze(0))\n",
        "print(out.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "torch.Size([1, 1000])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vqszlxEE0Bk0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 346
        },
        "outputId": "a698b34d-00a8-4067-ddc5-180cb4c8eeaa"
      },
      "source": [
        "# 4. mismatched tensor type\n",
        "import torch.nn as nn\n",
        "L = nn.CrossEntropyLoss()\n",
        "outs = torch.randn(5,5)\n",
        "labels = torch.Tensor([1,2,3,4,0])\n",
        "lossval = L(outs,labels) # Calculate CrossEntropyLoss between outs and labels"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "RuntimeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-10-60a5d1aad216>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mlabels\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m3\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m4\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\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----> 6\u001b[0;31m \u001b[0mlossval\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mL\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mlabels\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# Calculate CrossEntropyLoss between outs and labels\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    725\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    726\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--> 727\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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    728\u001b[0m         for hook in itertools.chain(\n\u001b[1;32m    729\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\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/python3.7/dist-packages/torch/nn/modules/loss.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input, target)\u001b[0m\n\u001b[1;32m    960\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    961\u001b[0m         return F.cross_entropy(input, target, weight=self.weight,\n\u001b[0;32m--> 962\u001b[0;31m                                ignore_index=self.ignore_index, reduction=self.reduction)\n\u001b[0m\u001b[1;32m    963\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    964\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py\u001b[0m in \u001b[0;36mcross_entropy\u001b[0;34m(input, target, weight, size_average, ignore_index, reduce, reduction)\u001b[0m\n\u001b[1;32m   2466\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0msize_average\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mreduce\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[1;32m   2467\u001b[0m         \u001b[0mreduction\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_Reduction\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlegacy_get_string\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msize_average\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreduce\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2468\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mnll_loss\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlog_softmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mignore_index\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreduction\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   2469\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2470\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py\u001b[0m in \u001b[0;36mnll_loss\u001b[0;34m(input, target, weight, size_average, ignore_index, reduce, reduction)\u001b[0m\n\u001b[1;32m   2262\u001b[0m                          .format(input.size(0), target.size(0)))\n\u001b[1;32m   2263\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mdim\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m2\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2264\u001b[0;31m         \u001b[0mret\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_C\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_nn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnll_loss\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_Reduction\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_enum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreduction\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mignore_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   2265\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mdim\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m4\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2266\u001b[0m         \u001b[0mret\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_C\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_nn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnll_loss2d\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_Reduction\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_enum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreduction\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mignore_index\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mRuntimeError\u001b[0m: expected scalar type Long but found Float"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CZwgwup_1dgS",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "aaf1de76-7ef2-4ca4-b87d-8482a3117249"
      },
      "source": [
        "# 4. mismatched tensor type (fixed)\n",
        "labels = labels.long()\n",
        "lossval = L(outs,labels)\n",
        "print(lossval)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor(2.6215)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dSuNdA8F06dK"
      },
      "source": [
        "**3. More on dataset and dataloader**\r\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "in84z_xu1rE6"
      },
      "source": [
        "A dataset is a cluster of data in a organized way. A dataloader is a loader which can iterate through the data set."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "34zfh-c22Qqs"
      },
      "source": [
        "Let a dataset be the English alphabets \"abcdefghijklmnopqrstuvwxyz\""
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TaiHofty1qKA"
      },
      "source": [
        "dataset = \"abcdefghijklmnopqrstuvwxyz\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h0jwhVa12h3a"
      },
      "source": [
        "A simple dataloader could be implemented with the python code \"for\""
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bWC5Wwbv2egy"
      },
      "source": [
        "for datapoint in dataset:\r\n",
        "  print(datapoint)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n33VKzkG2y2U"
      },
      "source": [
        "When using the dataloader, we often like to shuffle the data. This is where torch.utils.data.DataLoader comes in handy. If each data is an index (0,1,2...) from the view of torch.utils.data.DataLoader, shuffling can simply be done by shuffling an index array. \r\n",
        "\r\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9MXUUKQ65APf"
      },
      "source": [
        "torch.utils.data.DataLoader will need two imformation to fulfill its role. First, it needs to know the length of the data. Second, once torch.utils.data.DataLoader outputs the index of the shuffling results, the dataset needs to return the corresponding data."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BV5txsjK5j4j"
      },
      "source": [
        "Therefore, torch.utils.data.Dataset provides the imformation by two functions, `__len__()` and `__getitem__()` to support torch.utils.data.Dataloader"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A0IEkemJ5ajD"
      },
      "source": [
        "import torch\r\n",
        "import torch.utils.data \r\n",
        "class ExampleDataset(torch.utils.data.Dataset):\r\n",
        "  def __init__(self):\r\n",
        "    self.data = \"abcdefghijklmnopqrstuvwxyz\"\r\n",
        "  \r\n",
        "  def __getitem__(self,idx): # if the index is idx, what will be the data?\r\n",
        "    return self.data[idx]\r\n",
        "  \r\n",
        "  def __len__(self): # What is the length of the dataset\r\n",
        "    return len(self.data)\r\n",
        "\r\n",
        "dataset1 = ExampleDataset() # create the dataset\r\n",
        "dataloader = torch.utils.data.DataLoader(dataset = dataset1,shuffle = True,batch_size = 1)\r\n",
        "for datapoint in dataloader:\r\n",
        "  print(datapoint)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nTt-ZTid9S2n"
      },
      "source": [
        "A simple data augmentation technique can be done by changing the code in `__len__()` and `__getitem__()`. Suppose we want to double the length of the dataset by adding in the uppercase letters, using only the lowercase dataset, you can change the dataset to the following."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7Wn3BA2j-NXl"
      },
      "source": [
        "import torch.utils.data \r\n",
        "class ExampleDataset(torch.utils.data.Dataset):\r\n",
        "  def __init__(self):\r\n",
        "    self.data = \"abcdefghijklmnopqrstuvwxyz\"\r\n",
        "  \r\n",
        "  def __getitem__(self,idx): # if the index is idx, what will be the data?\r\n",
        "    if idx >= len(self.data): # if the index >= 26, return upper case letter\r\n",
        "      return self.data[idx%26].upper()\r\n",
        "    else: # if the index < 26, return lower case, return lower case letter\r\n",
        "      return self.data[idx]\r\n",
        "  \r\n",
        "  def __len__(self): # What is the length of the dataset\r\n",
        "    return 2 * len(self.data) # The length is now twice as large\r\n",
        "\r\n",
        "dataset1 = ExampleDataset() # create the dataset\r\n",
        "dataloader = torch.utils.data.DataLoader(dataset = dataset1,shuffle = True,batch_size = 1)\r\n",
        "for datapoint in dataloader:\r\n",
        "  print(datapoint)"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}