Datasets:
Tasks:
Text Generation
Sub-tasks:
language-modeling
Languages:
Japanese
Multilinguality:
monolingual
Size Categories:
10K<n<100K
Source Datasets:
SkelterLabsInc/JaQuAD
ArXiv:
License:
Update qg_jaquad.py
Browse files- qg_jaquad.py +1 -1
qg_jaquad.py
CHANGED
|
@@ -6,7 +6,7 @@ _DESCRIPTION = """
|
|
| 6 |
[JaQuAD](https://github.com/SkelterLabsInc/JaQuAD) dataset for question generation (QG) task. The test set of the original
|
| 7 |
data is not publicly released, so we randomly sampled test questions from the training set.
|
| 8 |
"""
|
| 9 |
-
_URL = 'https://huggingface.co/datasets/
|
| 10 |
_URLS = {
|
| 11 |
'train': ['{}/train{:02d}.jsonl'.format(_URL, i) for i in range(28)],
|
| 12 |
'test': ['{}/test{:02d}.jsonl'.format(_URL, i) for i in range(4)],
|
| 6 |
[JaQuAD](https://github.com/SkelterLabsInc/JaQuAD) dataset for question generation (QG) task. The test set of the original
|
| 7 |
data is not publicly released, so we randomly sampled test questions from the training set.
|
| 8 |
"""
|
| 9 |
+
_URL = 'https://huggingface.co/datasets/lmqg/qg_jaquad/raw/main/data/processed'
|
| 10 |
_URLS = {
|
| 11 |
'train': ['{}/train{:02d}.jsonl'.format(_URL, i) for i in range(28)],
|
| 12 |
'test': ['{}/test{:02d}.jsonl'.format(_URL, i) for i in range(4)],
|