深度学习

# -*- coding: utf-8 -*-
"""
Created on Wed Sep 15 17:15:50 2021

@author: 11651
"""
import os
os.environ["KMP_DUPLICATE_LIB_OK"]="TRUE"

import torch
import torchvision
from torch.utils import data
from torchvision import transforms
import matplotlib.pyplot as plt


# 通过ToTensor实例将图像数据从PIL类型变换成32位浮点数格式
# 并除以255使得所有像素的数值均在0到1之间
trans = transforms.ToTensor()
mnist_train = torchvision.datasets.FashionMNIST(
    root="./data", train=True, transform=trans, download=True)
mnist_test = torchvision.datasets.FashionMNIST(
    root="./data", train=False, transform=trans, download=True)

len(mnist_train), len(mnist_test)

mnist_train[0][0].shape

#生成批次数据
data_loader_train = torch.utils.data.DataLoader(dataset=mnist_train,
                                                batch_size = 18,
                                                shuffle = True)

data_loader_test = torch.utils.data.DataLoader(dataset=mnist_test,
                                               batch_size = 18,
                                               shuffle = True)

def get_fashion_mnist_labels(labels):  #@save
    """返回Fashion-MNIST数据集的文本标签。"""
    text_labels = ['t-shirt', 'trouser', 'pullover', 'dress', 'coat',
                   'sandal', 'shirt', 'sneaker', 'bag', 'ankle boot']
    return [text_labels[int(i)] for i in labels]

def show_images(imgs, num_rows, num_cols, titles=None, scale=1.5):  #@save
    """Plot a list of images."""
    figsize = (num_cols * scale, num_rows * scale)
    _, axes = plt.subplots(num_rows, num_cols, figsize=figsize)
    axes = axes.flatten()
    for i, (ax, img) in enumerate(zip(axes, imgs)):
        if torch.is_tensor(img):
            # 图片张量
            ax.imshow(img.numpy(), cmap='gray')
        else:
            # PIL图片
            ax.imshow(img)
        ax.axes.get_xaxis().set_visible(False)
        ax.axes.get_yaxis().set_visible(False)
        if titles:
            ax.set_title(titles[i])
    return axes

X, y = next(iter(data_loader_train))
show_images(X.reshape(18, 28, 28), 2, 9, titles=get_fashion_mnist_labels(y))