FewCLUE: OCNLI-FC (Natural Language Inference)

FewCLUE's few-shot cut of OCNLI: classify a Chinese premise-hypothesis pair as entailment, neutral or contradiction, learned from 32 labelled training pairs.

unassessed

This page is a discovery lead. Nobody has yet assessed it against the catalogue contract, so it carries no disposition. Absence of evidence here is not evidence of staleness.
Categoryreasoning
Subcategorythree-way natural language inference over native Chinese sentence pairs, few-shot cut of OCNLI
Page statusunknown
Metricaccuracy
Directionhigher_is_better
Unit%
Dataset size2520
PublisherCLUE team

What it measures

A premise and hypothesis sentence pair, natively authored in Chinese across five genres; the model classifies their relationship as entailment, neutral or contradiction, learned few-shot from 32 labelled training pairs -- the few-shot cut of the CLUE benchmark's OCNLI task.

Task format

Three-way natural language inference (entailment / neutral / contradiction), graded on the single correct label; evaluated from a 32-example few-shot training split, one of five parallel splits FewCLUE provides for this task.

Models reporting this benchmark

No model card in ModelSpec reports this benchmark yet.

Data

This page as JSON · Edit on GitHub