n-taku/CIFIR10_Sample.ipynb

## CIFIR10_Sample.ipynb
{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CIFIR10_Sample.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "wxbgeF4jI58s",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 317
        },
        "outputId": "d910555d-6975-4455-fcd2-1682443220df"
      },
      "source": [
        "import torch\n",
        "import torchvision\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "transform = torchvision.transforms.Compose([torchvision.transforms.Resize((32,32)),torchvision.transforms.ToTensor(),torchvision.transforms.Normalize((0.5,), (0.5,))])\n",
        "trainset = torchvision.datasets.CIFAR10(root = \"Dataset\", train = True, download = True, transform = transform)\n",
        "\n",
        "# 最初の画像(カエル)を表示\n",
        "image = trainset[0][0].clone().detach().numpy()\n",
        "image = image.transpose(1,2,0)\n",
        "image = image * np.array((0.5,0.5,0.5)) + np.array((0.5,0.5,0.5))\n",
        "image = image.clip(0,1)\n",
        "plt.imshow(image)\n",
        "\n",
        "# ローダーのshape\n",
        "trainloader = torch.utils.data.DataLoader(trainset, batch_size = 50, shuffle = True, num_workers = 2)\n",
        "print(iter(trainloader).next()[0].shape)\n",
        "print(iter(trainloader).next()[1].shape)\n"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Files already downloaded and verified\n",
            "torch.Size([50, 3, 32, 32])\n",
            "torch.Size([50])\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD5CAYAAADhukOtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nE1oNbl/zewAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}
	{
	"nbformat": 4,
	"nbformat_minor": 0,
	"metadata": {
	"colab": {
	"name": "CIFIR10_Sample.ipynb",
	"provenance": [],
	"collapsed_sections": []
	},
	"kernelspec": {
	"name": "python3",
	"display_name": "Python 3"
	}
	},
	"cells": [
	{
	"cell_type": "code",
	"metadata": {
	"id": "wxbgeF4jI58s",
	"colab_type": "code",
	"colab": {
	"base_uri": "https://localhost:8080/",
	"height": 317
	},
	"outputId": "d910555d-6975-4455-fcd2-1682443220df"
	},
	"source": [
	"import torch\n",
	"import torchvision\n",
	"import matplotlib.pyplot as plt\n",
	"import numpy as np\n",
	"\n",
	"transform = torchvision.transforms.Compose([torchvision.transforms.Resize((32,32)),torchvision.transforms.ToTensor(),torchvision.transforms.Normalize((0.5,), (0.5,))])\n",
	"trainset = torchvision.datasets.CIFAR10(root = \"Dataset\", train = True, download = True, transform = transform)\n",
	"\n",
	"# 最初の画像(カエル)を表示\n",
	"image = trainset[0][0].clone().detach().numpy()\n",
	"image = image.transpose(1,2,0)\n",
	"image = image * np.array((0.5,0.5,0.5)) + np.array((0.5,0.5,0.5))\n",
	"image = image.clip(0,1)\n",
	"plt.imshow(image)\n",
	"\n",
	"# ローダーのshape\n",
	"trainloader = torch.utils.data.DataLoader(trainset, batch_size = 50, shuffle = True, num_workers = 2)\n",
	"print(iter(trainloader).next()[0].shape)\n",
	"print(iter(trainloader).next()[1].shape)\n"
	],
	"execution_count": 7,
	"outputs": [
	{
	"output_type": "stream",
	"text": [
	"Files already downloaded and verified\n",
	"torch.Size([50, 3, 32, 32])\n",
	"torch.Size([50])\n"
	],
	"name": "stdout"
	},
	{
	"output_type": "display_data",
	"data": {
	"image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD5CAYAAADhukOtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nE1oNbl/zewAAAABJRU5ErkJggg==\n",
	"text/plain": [
	"<Figure size 432x288 with 1 Axes>"
	]
	},
	"metadata": {
	"tags": []
	}
	}
	]
	}
	]
	}