用mxnet实现cnn图像分类

import numpy as np
import os, time, shutil

from mxnet import gluon, image, init, nd
from mxnet import autograd as ag
from mxnet.gluon import nn
from mxnet.gluon.data.vision import transforms
from gluoncv.utils import makedirs
from gluoncv.model_zoo import get_model

We set the hyperparameters as following:

save_frequency = 1

模型保存路径

save_dir = ‘./save_model’

类别数量

classes = 3

全部数据集迭代次数

epochs = 30
lr = 0.001

如果有多个gpu可以用

per_device_batch_size = 2

优化器参数

momentum = 0.9

权值衰减参数，正则作用，一抑制过拟合

wd = 0.0001

学习率调节参数

lr_factor = 0.75
lr_steps &#