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 cased cnn daily mail dataset #25

Open
ruili33 opened this issue Nov 1, 2022 · 0 comments
Open

About cased cnn daily mail dataset #25

ruili33 opened this issue Nov 1, 2022 · 0 comments

Comments

@ruili33
Copy link

ruili33 commented Nov 1, 2022

Thank you for sharing the great work. Are there any hyper-parameters that should be changed when using the cased dataset? I only changed the dataset and the pre-trained model, but got even poorer results compared to the uncased model. By the way, I'm wondering why the article part in the cased data is still uncased like the photo below.
截屏2022-11-01 10 11 13

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