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.

Sign up for GitHub

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

Jump to bottom

Did you train BART with the input sequence length up to 2048 tokens? #13

Open
zhangyao1994 opened this issue Sep 28, 2022 · 0 comments
Open

Did you train BART with the input sequence length up to 2048 tokens? #13

zhangyao1994 opened this issue Sep 28, 2022 · 0 comments

Comments

@zhangyao1994
Copy link

Hi, in your QMSum paper, the 5.1 Implementation Details section mentioned that you used fairseq library to implement BART model and you truncate the input text to 2048 tokens.
Did you increase the max sequence length to 2048 for the BART model? May I ask how did you implement that?
I saw some suggestions in the following issue and wonder how exactly you implemented your finetuning.
facebookresearch/fairseq#1685
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@zhangyao1994