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

相同的模型结构,不同的参数,预测时间差异很大 #2531

Open
terrencew opened this issue Oct 10, 2024 · 2 comments
Open

相同的模型结构,不同的参数,预测时间差异很大 #2531

terrencew opened this issue Oct 10, 2024 · 2 comments
Assignees

Comments

@terrencew
Copy link

terrencew commented Oct 10, 2024

私有训练了一个OCR识别模型,训出来两个版本,模型结构一模一样,保存到硬盘上大小也没有区别。

现象是:
fastdeploy输入paddle模型,同样的代码,
选择onnx cpu推理,thread=1,先后进行了多次测试,速度测试稳定。
但结果差异巨大,一个是10s,一个是20s,区别只是识别模型用了一个结构的不同版本,也就是网络结构一致,weights值或者训练时的epoch不同。

请问大家遇到过这种问题吗?

@terrencew
Copy link
Author

terrencew commented Oct 10, 2024

补充:Paddle后端GPU推理两者都一样, Onnx后端GPU速度也是后者慢前者快,大概在4vs8左右

@terrencew
Copy link
Author

Repo分支信息
https://github.com/PaddlePaddle/FastDeploy/tree/develop
Commit ID cd0ee79

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

2 participants