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

Are codes for pretraining available? #10

Closed
PrettyMeng opened this issue Aug 3, 2020 · 3 comments
Closed

Are codes for pretraining available? #10

PrettyMeng opened this issue Aug 3, 2020 · 3 comments

Comments

@PrettyMeng
Copy link

It seems that this repository only contains the code to perform finetuning pretrained RoBERTa. Are code for pretraining available now? Can you possibly add some command example for doing TAPT? Any advice or explanation will be highly appreciated. Thanks in advance!

@PrettyMeng PrettyMeng changed the title Are code for pretraining available? Are codes for pretraining available? Aug 3, 2020
@kernelmachine
Copy link
Contributor

Hi there, check the ADAPTIVE_PRETRAINING.mdfile for DAPT/TAPT commands

@PrettyMeng
Copy link
Author

Hi there, check the ADAPTIVE_PRETRAINING.mdfile for DAPT/TAPT commands

Thanks for your quick reply! But just to clarify, in my case, the script is not working when transformers==2.4.1 as in environment.yml. In my case, it works well when transformers==2.8.0.

@shizhediao
Copy link

shizhediao commented Aug 23, 2020

Hi there, check the ADAPTIVE_PRETRAINING.mdfile for DAPT/TAPT commands

Thanks for your quick reply! But just to clarify, in my case, the script is not working when transformers==2.4.1 as in environment.yml. In my case, it works well when transformers==2.8.0.

Yes, same here.
When I am trying to run DAPT, there is an error ""ImportError: cannot import name 'MODEL_WITH_LM_HEAD_MAPPING""
I need to upgrade the transformers to 2.8.0 and it works well.
@kernelmachine

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

3 participants