python main.py
在ResNet18上跑50个epochs准确率:91%,但是太慢了,所以换LetNet5,来验证hyper-parameter的影响
在LeNet5上跑准确率:
epochs:20: 68.3%,batch_size:64
epochs:20: 66.5%,batch_size:50
epochs:10: 63.1%,batch_size:50
结论:epochs增加时准确率会上升,但是epochs过多会导致过拟合,且准确率越往后增长越慢,当前epochs还算少。batch_size变大时,训练数据变多,效果变好
-
Notifications
You must be signed in to change notification settings - Fork 0
l-wd/AiSafety-CIFAR10
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
No description, website, or topics provided.
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published