C-Eval | GPTtopic

C-Eval is a multi-level and multidisciplinary Chinese assessment kit suitable for large language models, jointly launched by researchers from Shanghai Jiao Tong University, Tsinghua University, and the University of Edinburgh in May 2023. It includes 13948 multiple-choice questions, covering 52 different disciplines and four difficulty levels, to evaluate the Chinese understanding ability of large language models.

data statistics

Relevant Navigation

FlagEval

FlagEval, jointly developed by Zhiyuan Research Institute and multiple university teams

PubMedQA

PubMedQA is a biomedical research question and answer dataset that includes 1K expert annotated, 61.2K unlabeled, and 211.3K manually generated QA instances. The ranking currently includes medical test scores for 18 models.

LLMEval3

LLMEval is a large-scale model evaluation benchmark launched by the NLP Laboratory of Fudan University.

SuperCLUE

SuperCLUE is a comprehensive evaluation benchmark for Chinese general large models, which evaluates the capabilities of models from three different dimensions: basic ability, professional ability, and Chinese characteristic ability.

HELM

HELM, also known as Holistic Evaluation of Language Models, is a large-scale model evaluation system developed by Stanford University.

MMLU

Large scale multitasking language comprehension benchmark

No comments

No comments...