Datasets:

tau
/

scrolls

Tasks:

Question Answering

Text Generation

Sub-tasks:

multiple-choice-qa

natural-language-inference

Languages:

ArXiv:

Tags:

query-based-summarization

Dataset card Files Files and versions

eladsegal commited on Dec 5, 2021

Commit

451ff2c

·

1 Parent(s): fcc2240

Update metrics/bleu.py

Files changed (1) hide show

metrics/bleu.py +1 -1

metrics/bleu.py CHANGED Viewed

@@ -117,6 +117,6 @@ def compute_bleu(reference_corpus, translation_corpus, max_order=4, smooth=False
     }
-def bleu_postprocess_text(text):
     # TODO: Tokenize properly
     return text.split()

     }
+def postprocess_text(text):
     # TODO: Tokenize properly
     return text.split()