-
Notifications
You must be signed in to change notification settings - Fork 20
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
add subtsring decontamination #16
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Great thank you for adding this! I just left a comment on a missing file from which you import humaneval reader. It would also be helpful if you can add a readme with what is needed to run the code.
@ChenghaoMou we can probably add instructions to execute your code too in that readme (forgot to mention it in your PR)
from tqdm import tqdm | ||
from multiprocessing import Pool | ||
|
||
from human_eval.data import read_problems |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think the file human_eval
is missing (btw we also have HumanEval on datasets for the future :) )
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Oh actually I directly installed https://github.com/openai/human-eval as a package. I could modify the code to use HF datasets instead yes.
assert len(data) == 500 | ||
|
||
# Checksum / version issues here | ||
# dataset = load_dataset("mbpp", split="test") |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I also noticed that, you will need to upgrade datasets version (if the error still happens there is the ignore_verifications=True
argument)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
thanks for adding datasets
Exact-substring match for decontamination #13
This removes 336 files from the
python-permissive
dataset and 292 from thepython-permissive-dedup
dataset. All these removals match HumanEval samples.