Language Accuracy
收藏arXiv2025-09-30 收录
下载链接：
https://huggingface.co/datasets/airesearch/WangchanThaiInstruct
下载链接
链接失效反馈官方服务：
资源简介：
该数据集对模型响应进行了基于代码切换的指标评估，以检验其在自然语言习惯方面的表现。此外，该评估是通过使用500个样本指令来进行的。任务目标是进行语言评估。
This dataset conducts metric evaluation on model responses based on code-switching to assess the model's performance in adhering to natural language conventions. Furthermore, this evaluation is implemented using 500 sample instructions. The task objective of this work is language assessment.
提供机构：
airesearch
