deca-ai/open-synth-battles
收藏Hugging Face2025-09-03 更新2025-09-13 收录
下载链接：
https://hf-mirror.com/datasets/deca-ai/open-synth-battles
下载链接
链接失效反馈官方服务：
资源简介：
OpenSynth Battles是一个基准测试数据集，其中五种语言模型通过生成对同一提示的回应进行竞争。这些模型包括gpt-oss-120b、deepseek-v3.1-thinking、deepseek-v3.1-instruct、moonshotai/kimi-k2-instruct和deepseek-r1-0528。每个提示都与所有五种模型的回应配对，并由gpt-oss-120b模型作为自动化裁判进行评估和打分。
OpenSynth Battles is a benchmarking dataset where five language models compete by generating responses to the same prompt. The models include gpt-oss-120b, deepseek-v3.1-thinking, deepseek-v3.1-instruct, moonshotai/kimi-k2-instruct, and deepseek-r1-0528. Each prompt is paired with responses from all five models, and their outputs are evaluated and scored by the gpt-oss-120b model acting as an automated judge.
提供机构：
deca-ai
