26 datasets found

The Super-NaturalInstructions (SNI) dataset is a collection of 1761 diverse NLP tasks belonging to one of 76 task types.

A reading comprehension dataset of 32,091 document-question-answer triplets that are true or false.

The WSC dataset is a benchmark for coreference resolution tasks.

The CB dataset is a benchmark for textual entailment tasks.

The BoolQ dataset is a context-based yes-no question answering dataset.

The dataset used in the paper is the SuperGLUE benchmark, which includes 17 tasks: STS-B, MRPC, MNLI, QNL, QNLI, CoLA, SST-2, MRPC, GLUE, NLI, NQ, ReCoRD, ReCoRD-Sub,...

Chinanews dataset

Online shopping review dataset

Chinese sentiment classification and entailment models

The ReCoRD dataset is a benchmark for reading comprehension tasks.

The dataset consists of 544 documents sourced from the BBC Sport website.

The dataset consists of newswire articles available in five languages.

RTE dataset

The BEiT dataset used for the experiments in the paper.

The 20Newsgroups data set is a dataset of 18,846 instances of newsgroup documents.

The 20-Newsgroups dataset is a collection of text documents.

The dataset used for the experiments on Twitter and Pinterest.

The dataset used for the experiments across ten models– ranging from bag-of-words models to pre-trained transformers– and find that a model having higher AUC does not necessarily...

The dataset used in the paper is not explicitly described. However, it is mentioned that the authors used various benchmarks such as Wiki-40B, PG-19, C4, etc.

The dataset used in the paper is a language domain dataset, specifically for sentiment classification, named AG News. The dataset is used to evaluate the performance of...