Getting Started with PyTorch Deep Learning in Six Days (1/6)

Get started with PyTorch deep learning in six days (1/6)

First introduction to PyTorch

Pytorch deep learning quick start simple tutorial, suitable for all novices to learn and lay the foundation of the framework.
Follow my pace and learn step by step, you can master it in a week

Follow my pace and learn step by step, you can master it in a week

import torch    #导入torch库，PyTorch是一个基于Torch的Python开源机器学习库

#查看当前机器的torch版本和cuda版本
print(torch.__version__)

#查看当前机器的gpu是否可用，可用即为True
torch.cuda.is_available()

dir(): View the subfunctions under the function

dir(torch.nn)

help(): View function usage

help(torch.nn.Softmax)

from torch.utils.data import Dataset
import os   #处理图像
from PIL import Image   #打开图像

Dataset(): Get data and its labels

class Mydata(Dataset):
    def __init__(self,root_dir,label_dir):     #初始化：根据类创建实例时首先要运行的函数，为后面的函数提供全局变量self
        self.root_dir=root_dir
        self.label_dir=label_dir
        self.path=os.path.join(root_dir,label_dir)     #拼接地址
        self.img_path=os.listdir(self.path)   #该地址下的文件名转化为列表形式

    def __getitem__(self, idx):     #idx:图像的索引序号
        img_name=self.img_path[idx]
        img_item_path=os.path.join(self.path,img_name)
        img=Image.open(img_item_path)
        label=self.label_dir
        return img,label

    def __len__(self):
        return (len(self.img_path))

Instantiate

root_dir="../datasets/hymenoptera_data/train"
label_dir="ants"
ants_dataset=Mydata(root_dir,label_dir)
img,label=ants_dataset[1]
img,label