Jamp(tomo-vv/temporalNLI_dataset) is the Japanese temporal inference benchmark. This dataset consists of templates, test data, and training data. Template subset containing template, time format, or time span in their names are split based on tense fragment, time format, or time span, respectively.

	Dataset Details




	


	Dataset Sources… See the full description on the dataset page: https://huggingface.co/datasets/zenless-lab/jamp.