LV-Eval

What is LV-Eval?

LV-Eval is a challenging long-context benchmark with five length levels (16k, 32k, 64k, 128k, and 256k) reaching up to 256k words. The average number of words…

Released

2024-02-18

Evaluates

长文本, long-context, 大语言模型, LLM, 长上下文, Long Context, 合作共建, Co-Built, 不支持, Unsupported

Openness

restricted

Importer

OpenCompass Hub

Reported scores

0

Source provenance

• Original evidence https://hub.opencompass.org.cn/dataset-detail/LV-Eval

LV-Eval paper and code

• Paper https://arxiv.org/abs/2402.05136
• Code https://github.com/infinigence/LVEval

Which sources cite LV-Eval?

• LV-Eval

Related 长文本, long-context, 大语言模型, LLM, 长上下文, Long Context, 合作共建, Co-Built, 不支持, Unsupported benchmarks

• A-Bench
• A-OKVQA
• AA-LCR
• AbsPyramid
• ActionAtlas
• Ada-LEval
• AECBench
• AFQMC