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

Details contrastive training #3

Open
boscoj2008 opened this issue Jan 17, 2023 · 1 comment
Open

Details contrastive training #3

boscoj2008 opened this issue Jan 17, 2023 · 1 comment

Comments

@boscoj2008
Copy link

boscoj2008 commented Jan 17, 2023

During contrastive pre-training, do you also set batch_size to 1024? or is it the default 8? Thanks

@rpeeters85
Copy link
Contributor

The batch size to be used during pre-training can be set manually (this is the batch size you define in the command line). The value used in the paper is 1024. On the other hand, the batch size for the fine-tuning step is set to 64, you can find this in the .sh script of the corresponding fine-tuning e.g. src/contrastive/lspc/run_finetune_siamese_frozen_roberta.sh and change it there if you want to change the batch size used for fine-tuning.

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