[python]自定义数据集多分类ResNet-18

最近需要训练关于自定义数据集的ResNet18model

1. 自定义数据集:　用到了数据增强，摄像头抓拍，图片处理（标签，大小名称）等
2. 数据处理好后，需要处理input , 重写class Mydataset
3. import os
   from matplotlib import image
   import numpy as np
   import torch
   import torch.nn as nn
   import matplotlib.image as image
   from torch.utils.data import DataLoader, Dataset
   from torchvision import models, transforms
   import torchvision.models as models
   from sklearn.model_selection import train_test_split
   import matplotlib.pyplot as plt
   from PIL import Image
   from torch.autograd import Variable
   from torchvision.models.resnet import resnet18
   from torch.optim.lr_scheduler import *

class MyDataSet(Dataset):
‘’’
定义数据集，用于将读取到的图片数据转换并处理成CNN神经网络需要的格式
‘’’
def init(self,root,transfrom=None,train=True,test=False):
self.test = test
self.train = trai