257 datasets found

The DREAM dataset is a dialogue-based multiple-choice QA dataset, introduced by Sun et al. (2019). It was collected from English-as-a-foreign-language examinations, designed by...

Dialogue summarization datasets including SAMSum, DialogSum, TODSum, and DREAM.

The XTREME-UP dataset is a cross-lingual question answering task that asks a model to predict the correct English answer span given a non-English question and an English answer...

The XOR-ATTRIQA dataset is a classification task where model is asked to predict whether the provided answer to the question is supported by the given passage context, which...

The Stanford Human Preferences (SHP) dataset is sourced from Reddit with various subreddits that focus on QA. Preferences have been extracted from the accumulated up- and...

The Pile dataset consists of 800GB text from 22 domains. Cynical selection naturally prefers text data based on the target corpus.

FigureQA is a dataset for visual question answering, containing line plots, bar charts, pie plots, and dot line plots.

The UNK-VQA dataset is a dataset for evaluating the ability of large language models to answer questions when the answer is unknown.

This dataset contains question-answer pairs for training a semantic parser.

This dataset contains question-answer pairs for training a semantic parser.

The dataset is used for concept graph recovery and question answering in NLP education.

A dataset for question answering, containing Wikipedia graph for question answering.

A benchmark for question answering research is introduced, which includes a large dataset of natural questions.

A video understanding dataset of videos recorded in outside settings, including video question answering and video evidence selection.

WebSRC dataset for web-based structural reading comprehension.

Topological Information Enhanced model (TIE) for web-based structural reading comprehension on web pages.

Educational Question Generation (QG) dataset, used to train and evaluate QG models.

Walking down the memory maze: beyond context limit through interactive reading.

WikiWeb2M is a page-level multimodal Wikipedia dataset.

Document structure in long document transformers.