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

About data flipping #101

Open
zzzendurance opened this issue Jul 1, 2024 · 5 comments
Open

About data flipping #101

zzzendurance opened this issue Jul 1, 2024 · 5 comments

Comments

@zzzendurance
Copy link

Thank you for reading my question. In the code, the data is flipped in the last dimension: xz.flip([-1]), so I want to ask if it can be changed in other dimensions (you only need to answer me this question), because I am worried that if I only change xz.flip([-1]), the code logic elsewhere is related to this code logic.

@zzzendurance
Copy link
Author

I have just tried to flip another dimension and found that it works successfully, but I am not sure if this change is effective. If the author sees my question, thank you for your answer

@Karn3003
Copy link

Karn3003 commented Nov 10, 2024

Hi @zzzendurance , It should be 1 right? As I do think in out_b it should be out_b.flip([1]) instead of -1. As we need to flip the tokens not the dimensions across the tokens.

@zzzendurance
Copy link
Author

Hi @zzzendurance , It should be 1 right? As I do think in out_b it should be out_b.flip([1]) instead of -1. As we need to flip the tokens not the dimensions across the tokens.

I think what you said makes sense. I’ve been following this code for a long time, and I almost forgot about it. I just looked it up, and it should be this line of code. This line is likely flipping the c dimension in the (b, t, c) format, right?

@Karn3003
Copy link

Hi @zzzendurance , It should be 1 right? As I do think in out_b it should be out_b.flip([1]) instead of -1. As we need to flip the tokens not the dimensions across the tokens.

I think what you said makes sense. I’ve been following this code for a long time, and I almost forgot about it. I just looked it up, and it should be this line of code. This line is likely flipping the c dimension in the (b, t, c) format, right?

Yes, it's flipping the c dimension in (b, t, c). What kind of task did you perform, and did this change work successfully??

@zzzendurance
Copy link
Author

Hi @zzzendurance , It should be 1 right? As I do think in out_b it should be out_b.flip([1]) instead of -1. As we need to flip the tokens not the dimensions across the tokens.

I think what you said makes sense. I’ve been following this code for a long time, and I almost forgot about it. I just looked it up, and it should be this line of code. This line is likely flipping the c dimension in the (b, t, c) format, right?

Yes, it's flipping the c dimension in (b, t, c). What kind of task did you perform, and did this change work successfully??

My task is speech emotion recognition. I remember that this change worked successfully, but the effect did not seem to change much.

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