evaleval/every_eval_ever
收藏官方服务：
资源简介：
该数据集名为Every Eval Ever Dataset，主要收集了来自不同AI模型排行榜的评估结果。数据集包含多个排行榜的数据，如global_mmlu_lite、hfopenllm_v2和livecodebenchpro。每条数据记录包括模型信息（名称、ID、开发者）、排行榜名称、评估结果（以JSON字符串形式存储）以及其他用于确保可重复性的元数据。数据集通过GitHub Actions实现自动更新。
The dataset is named Every Eval Ever Dataset and primarily contains evaluation results from various AI model leaderboards. It includes data from multiple leaderboards (splits) such as global_mmlu_lite, hfopenllm_v2, and livecodebenchpro. Each data record includes model information (name, ID, developer), leaderboard name, evaluation results (stored as a JSON string), and additional metadata for reproducibility. The dataset is auto-updated via GitHub Actions.
提供机构：
evaleval
