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

[Bug report] Ineffective no_decay when using BERTAdam #33

Closed
xiaoda99 opened this issue Nov 18, 2018 · 1 comment
Closed

[Bug report] Ineffective no_decay when using BERTAdam #33

xiaoda99 opened this issue Nov 18, 2018 · 1 comment

Comments

@xiaoda99
Copy link
Contributor

https://github.com/huggingface/pytorch-pretrained-BERT/blob/master/examples/run_classifier.py#L505-L508

With this code, all parameters are decayed because the condition "parameter_name in no_decay" will never be satisfied.

I've made a PR #32 to fix it.
@xiaoda99 xiaoda99 changed the title [Bug report] Ineffective no_decay in BERTAdam [Bug report] Ineffective no_decay when using BERTAdam Nov 18, 2018
@thomwolf
Copy link
Member

You're right, thanks!

@maeotaku maeotaku mentioned this issue May 23, 2019
Closed
stevezheng23 added a commit to stevezheng23/transformers that referenced this issue Mar 24, 2020
@stevezheng23
Merge pull request huggingface#33 from stevezheng23/dev/zheng/coqa
d19ce97 
add coqa basic-runner as default coqa at-runner (cont.)
@Arij-Aladel Arij-Aladel mentioned this issue Jul 18, 2022
Closed
jameshennessytempus pushed a commit to jameshennessytempus/transformers that referenced this issue Jun 1, 2023
@jamesthesnake
Merge pull request huggingface#33 from jamesthesnake/pop
cef9b20 
Pop
@lwmlyy lwmlyy mentioned this issue Aug 15, 2023
Merged
1 task
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
2 participants
@xiaoda99 @thomwolf