Datasets:
Tasks:
Question Answering
Sub-tasks:
extractive-qa
Languages:
Basque
Size:
1K<n<10K
Tags:
dialogue-qa
License:
Dataset is broken
#4
by
hynky
HF staff
- opened
Hello I am unable to load the dataset here is the error:
Command: from datasets import load_dataset load_dataset("ixa-hitz/elkarhizketak", split="train", name="plain_text")
NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=1024378, num_examples=301, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='train', num_bytes=3104779, num_examples=1306, shard_lengths=None, dataset_name='elkarhizketak')}, {'expected': SplitInfo(name='validation', num_bytes=125667, num_examples=38, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='validation', num_bytes=373522, num_examples=161, shard_lengths=None, dataset_name='elkarhizketak')}, {'expected': SplitInfo(name='test', num_bytes=127640, num_examples=38, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='test', num_bytes=382877, num_examples=167, shard_lengths=None, dataset_name='elkarhizketak')}]