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

[Chatllama] facebook/opt-350m is missing in rlhf/model_list.py #323

Closed
tengxiaoliu opened this issue Apr 2, 2023 · 2 comments
Closed

[Chatllama] facebook/opt-350m is missing in rlhf/model_list.py #323

tengxiaoliu opened this issue Apr 2, 2023 · 2 comments

Comments

@tengxiaoliu
Copy link

Hi, thanks for the great repo!

In rlhf/model_list.py, "facebook/opt-350m" is missing in hf_models_causal_lm. Setting "model=facebook/opt-350m" in actor_config will cause "UnboundLocalError: local variable 'tokenizer' referenced before assignment" at rlhf/actor.py:176.

Or maybe simply add a warning for unsupported models when loading tokenizer at Line176 in chatllama/rlhf/actor.py.

@PierpaoloSorbellini
Copy link
Collaborator

Hi @tengxiaoliu, thanks for reaching out and for the advice.
We did miss the 350m version when adding the strings to model_list.py.
We have added opt-350m and an error message for unsupported models. (in PR #306)
Thanks again for the feedback!

@tengxiaoliu
Copy link
Author

Thanks for your reply! I'll close this issue.

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