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 reproduce the results of alpaca_eval #3

Open
jxs129 opened this issue Jan 13, 2025 · 4 comments
Open

how to reproduce the results of alpaca_eval #3

jxs129 opened this issue Jan 13, 2025 · 4 comments

Comments

@jxs129
Copy link

jxs129 commented Jan 13, 2025

Dear authors,

Thanks for the interesting paper!

I am trying to reproduce Table 4, some questions about AlpacaEval:

  • how to build the AlpacaEval dataset?
  • which script can reproduce the AlpacaEval results in Table 4?

thanks

@huangtiansheng
Copy link
Member

Hi,

Thanks for reaching out. Check out this script:

https://github.com/git-disl/Vaccine/blob/main/script/sft_finetune/alpaca.sh
https://github.com/git-disl/Vaccine/blob/main/script/vaccine_finetune/alpaca.sh

For initial experiment/reproduce, we don't recommend alpacaeval because it needs money and actually the benchmark itself is not that stable :)

@jxs129
Copy link
Author

jxs129 commented Jan 13, 2025

thanks for your response.

where can I find the data/alpaca.json?

@huangtiansheng
Copy link
Member

It is this file https://huggingface.co/datasets/tatsu-lab/alpaca_eval/blob/main/alpaca_eval_gpt4_baseline.json

Download and rename it to alpaca.json and put it to data directory. Then it should work.

@jxs129
Copy link
Author

jxs129 commented Jan 13, 2025

ok, great!

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