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

offline和online推理速度差异 #26

Open
CaA23187 opened this issue Jul 23, 2024 · 1 comment
Open

offline和online推理速度差异 #26

CaA23187 opened this issue Jul 23, 2024 · 1 comment

Comments

@CaA23187
Copy link

我在X86平台和ARM平台上测试offline和online的RTF, 发现online版本会慢上27倍左右. 但是从CPU占用率上看, 两个版本的GTCRN的使用率都接近100%, 请问这运行速度上的差异可能是什么原因导致的?

@Xiaobin-Rong
Copy link
Owner

online 版本比 offline 慢是合理的,考虑卷积的 online 和 offline 两种形式,前者实际上相当于写了一个 for 循环来处理数据,而后者直接对数据进行并行处理。如果考虑 RNN 的 online/offline 的话,虽然不能并行处理,但是后者的 pytorch 实现是用 C代码写的,所以还是会快。

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