Datasets:
lmqg
/

Modalities:
Text
Languages:
English
ArXiv:
Libraries:
Datasets
License:
asahi417 commited on
Commit
92b00a8
1 Parent(s): 13c7811
Files changed (1) hide show
  1. qg_squad.py +1 -0
qg_squad.py CHANGED
@@ -2,6 +2,7 @@ import datasets
2
  from datasets import load_dataset
3
  from datasets.tasks import Summarization
4
 
 
5
  _DESCRIPTION = """
6
  [SQuAD](https://rajpurkar.github.io/SQuAD-explorer/) evaluation set for the question generation (QG) models. The split
7
  of test and development set follows the ["Neural Question Generation"](https://arxiv.org/abs/1705.00106) work and is
 
2
  from datasets import load_dataset
3
  from datasets.tasks import Summarization
4
 
5
+ logger = datasets.logging.get_logger(__name__)
6
  _DESCRIPTION = """
7
  [SQuAD](https://rajpurkar.github.io/SQuAD-explorer/) evaluation set for the question generation (QG) models. The split
8
  of test and development set follows the ["Neural Question Generation"](https://arxiv.org/abs/1705.00106) work and is