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

How to use Beam Search for text generation? #619

Closed
cdj0311 opened this issue May 19, 2022 · 6 comments
Closed

How to use Beam Search for text generation? #619

cdj0311 opened this issue May 19, 2022 · 6 comments

Comments

@cdj0311
Copy link

cdj0311 commented May 19, 2022

Hi,
Thanks for this code. But I want know how to use Beam Search for text generation?

@StellaAthena
Copy link
Member

We do not currently support beam search for text generation. You will have to implement that yourself or use the WIP HuggingFace port

@cdj0311
Copy link
Author

cdj0311 commented May 26, 2022

We do not currently support beam search for text generation. You will have to implement that yourself or use the WIP HuggingFace port

Thanks, but have a script convert model to huggingface transformers load?

@StellaAthena
Copy link
Member

The model is on the EleutherAI HF page already.

@cdj0311
Copy link
Author

cdj0311 commented May 27, 2022

The model is on the EleutherAI HF page already.

enn, the 20B model on the HF page is too large to loaded, I trained a small model (with small.yml), how to convert it ?

@StellaAthena
Copy link
Member

@zphang where’s the function you wrote for this? I don’t see it in tools like I expected to.

@jason9693
Copy link
Member

jason9693 commented Sep 9, 2022

@cdj0311 you can use this script.
I've tested on 1B and 3B models and both of all is succeed.

#668

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