Datasets:

Modalities:
Text
Formats:
parquet
Languages:
English
ArXiv:
Libraries:
Datasets
Dask
License:

Where is the "download" button

#7
by damc - opened

I'm trying to download this dataset, but I can't see the "download" button, can anyone help?

You can download the splits in Parquet format at https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet. You get there by clicking the link "Auto-converted to Parquet"

Capture d’écran 2023-10-25 à 18.19.42.png

in the dataset viewer:

Capture d’écran 2023-10-25 à 18.20.36.png

Otherwise, you can look at the script to get the original data URLs, like: eg: https://s3.amazonaws.com/research.metamind.io/wikitext/wikitext-103-v1.zip

Neither the script not the original data URL are accessible.

Any update on this?

from datasets import load_dataset
testdata = load_dataset('Salesforce/wikitext', 'wikitext-2-raw-v1', split='test')

This is giving me BadRequestError

raise BadRequestError(message, response=response) from e
huggingface_hub.utils._errors.BadRequestError:  (Request ID: Root=1-687a2f83-18f216f33260191d6a44c627;41a27378-b782-4cde-bedc-77b5c94ade6d)

Bad request:
* Invalid input: expected array, received string * at paths * Invalid input: expected boolean, received string * at expand
✖ Invalid input: expected array, received string
  → at paths
✖ Invalid input: expected boolean, received string
  → at expand

Sign up or log in to comment