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

Can load Udop-Dual-Large-224, but not Udop-Unimodel-Large-224 #30

Closed
plamb-viso opened this issue Mar 2, 2023 · 4 comments
Closed

Can load Udop-Dual-Large-224, but not Udop-Unimodel-Large-224 #30

plamb-viso opened this issue Mar 2, 2023 · 4 comments

Comments

@plamb-viso
Copy link

I was able to load the dual-large tokenizer/config/model using the same method described in this comment

However, when I attempt to use the exact same code, but for udop-unimodel-large-224 I get:

Traceback (most recent call last):
    tok = UdopTokenizer.from_pretrained(
    return cls._from_pretrained(
    tokenizer = cls(*init_inputs, **init_kwargs)
    self.sp_model.Load(vocab_file)
    return self.LoadFromFile(model_file)
    return _sentencepiece.SentencePieceProcessor_LoadFromFile(self, arg)
RuntimeError: Internal: unk is not defined.
@zinengtang
Copy link
Collaborator

zinengtang commented Mar 2, 2023

Can I know what transformer and sentencepiece version do you use.

@plamb-viso
Copy link
Author

I am using the version from this PR which is based on the main branch and most recently 4.26.1

@zinengtang I realized that I can load the model/config from udop-unimodel-large-224 and the tokenizer from udop-dual-large-224 with no errors. Do you foresee this causing any problems or do the two tokenizers use the same settings?

@zinengtang
Copy link
Collaborator

different tokenizers shouldn't cause any issues, all configs are shared.

@zinengtang
Copy link
Collaborator

Our model dev uses 4.20.1. it might affect loading I think but I am not sure.

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