跳到内容

夜深了,注意休息,愿你今夜好梦。

AI 资源分类

大模型评测

共 11 条资源

当前分类链接 返回上级:大模型

HELM

HELM is a large model evaluation system introduced by Stanford University. The evaluation methodology consists of three main modules: scenarios, fitness, and metrics, and each evaluation run requires the specification of a scenario, a prompt to fit the model, and one or more metrics.

2026年4月15日 469 0 浏览 469,收藏 0
正文
强调色