Datasets:

stefanbschneider
/

lfqa-max-answer-length-512

Tasks:

Question Answering

Text2Text Generation

Modalities:

Formats:

Languages:

Size:

Tags:

Libraries:

License:

Dataset card Data Studio Files Files and versions Community

stefanbschneider commited on Feb 8

Commit

f47df80

·

verified ·

1 Parent(s): 5e48a8c

typo

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -56,7 +56,7 @@ Similar to ELI5 but with context.
 This dataset is a filtered version of [LLukas22/lfqa_preprocessed](https://huggingface.co/datasets/LLukas22/lfqa_preprocessed),
 which in turn is a processed and simplified version of of [vblagoje's](https://huggingface.co/vblagoje) *[lfqa_support_docs](https://huggingface.co/datasets/vblagoje/lfqa_support_docs)* and *[lfqa](https://huggingface.co/datasets/vblagoje/lfqa)* datasets.
-This dataset (stefanbschneider/lfqa_preprocessed) has filtered out overly long answers, based on the number of tokens in the answer using the LED tokenizer.
 It can be reproduced with the notebook `process-lfqa-dataset.ipynb`.
 LLukas22/lfqa_preprocessed             |  stefanbschneider/lfqa-max-answer-length-512

 This dataset is a filtered version of [LLukas22/lfqa_preprocessed](https://huggingface.co/datasets/LLukas22/lfqa_preprocessed),
 which in turn is a processed and simplified version of of [vblagoje's](https://huggingface.co/vblagoje) *[lfqa_support_docs](https://huggingface.co/datasets/vblagoje/lfqa_support_docs)* and *[lfqa](https://huggingface.co/datasets/vblagoje/lfqa)* datasets.
+I have filtered out overly long answers, based on the number of tokens in the answer using the LED tokenizer.
 It can be reproduced with the notebook `process-lfqa-dataset.ipynb`.
 LLukas22/lfqa_preprocessed             |  stefanbschneider/lfqa-max-answer-length-512