Datasets:

Languages:
English
Size Categories:
n>1T
ArXiv:
License:

Duplicate Key when Loading Dataset

#18
by Hhhhhao97 - opened

I got this error when loading the dataset

The examples at index 113, 152 have the key 8ea47781d1f36e5773e5f8b78576723052ca8dcc

Any solution on how to solve it?

Allen Institute for AI org

Hi, could you please provide more details on how you ran into this issue, like the commands you ran to reach this error?

After downloading the dataset, the error comes at loading the dataset

os.environ['DOLMA_DATA_DIR'] = 'dolma'
dataset = load_dataset("allenai/dolma", split="train")

FYI there are several repeated samples in this dataset and they're all assigned the same row['id'].
A temporary workaround: You can edit dolma.py to replace row['id'] as the key with a random key here and that will let you load it.

I confirm this issue. Getting exactly the same error.

Sign up or log in to comment