Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

YOLOv3模型训练好并保存后,再加载模型预测的输出shape变了(直接编译运行错误了) #106

Open
WorstCodeWay opened this issue Aug 2, 2020 · 0 comments

Comments

@WorstCodeWay
Copy link

把YOLOv3整个模型(模型和参数)训练好了保存之后,再加载之后预测的结果shape变了。不保存直接推断是3个元素的tensor。而现在是6个元素的tensor。如下:

对于一张图片的预测
训练完直接预测的输出(只有一个类别)shape
[1 52 52 3 6]
[1 26 26 3 6]
[1 13 13 3 6]

通过加载保存好的整个模型预测的输出(只有一个类别)shape
[1 52 52 18]
[1 26 26 3 6]
[1 13 13 18]
[1 52 52 3 6]
[1 26 26 3 6]
[1 13 13 18]

我使用的 Tensorflow2.2 GPU版本 Ubuntu16.04LTS

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant