ctu-aic/nli_it_collection
收藏Hugging Face2025-06-15 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/ctu-aic/nli_it_collection
下载链接
链接失效反馈官方服务：
资源简介：
这是一个包含捷克语和英语的自然语言推理（NLI）数据集集合，转换为基于FLAN方法的指令微调格式。数据集适用于对解码器LLM模型进行指令微调，包含了训练集、验证集和测试集。每个数据点包括前提、假设、答案等信息，并提供了对话格式，可直接用于指令微调训练。
This is a collection of Czech and English Natural Language Inference (NLI) datasets transformed into an instruction tuning format based on the FLAN approach. The dataset is suitable for instruction tuning of decoder LLM models and includes train, validation, and test sets. Each data point consists of premise, hypothesis, answer, and is provided in a conversational format ready for direct use in instruction tuning training.
提供机构：
ctu-aic
