m0_62722599的博客

1.总体流程2.代码①数据集划分import splitfolderssplitfolders.ratio(input='dataset', output='output', ratio=(0.8, 0.2))②全部流程import torchimport torch.nn as nnimport matplotlib.pyplot as plt # 用来画准确度和损失值import torch.nn.functional as F # 导入relu归一化函数，该.

1.代码import torch.nn as nnimport torchfrom torch.utils.data import DataLoaderimport torchvisionimport matplotlib.pyplot as pltfrom tqdm import tqdm# 神经网络模型class Net(nn.Module): def __init__(self): # 调用父类的构造函数 super(Net, self).