rl-bandits-lab/SEGALE-WMT24-Human-Eval
收藏Hugging Face2025-11-05 更新2025-11-15 收录
下载链接:
https://hf-mirror.com/datasets/rl-bandits-lab/SEGALE-WMT24-Human-Eval
下载链接
链接失效反馈官方服务:
资源简介:
SEGALE元评估数据集(人工标注)包含WMT24环境下长上下文机器翻译输出的专家人工错误注释。该数据集设计用于元评估、评估指标分析和奖励模型训练。
The SEGALE Meta Evaluation Dataset (Human-Annotated) contains expert human error annotations for long-context MT outputs in WMT24 settings. It is designed for meta-evaluation, analysis of evaluation metrics, and reward-model training.
提供机构:
rl-bandits-lab



