obalcells/raw-fact-extraction

该数据集包含多个配置，涉及不同的模型和任务，如Meta-Llama-3.1-8B-Instruct、gemma-2-9b-it、gemma-2b-it等。每个配置的特征包括数据集名称、子集、原始数据集索引、原始数据集分割、查询ID、完成索引、对话内容、完成内容以及事实信息等。数据集主要用于测试和评估模型的性能，特别是在对话生成、事实提取和问答任务中的表现。

This dataset contains multiple configurations involving different models and tasks, such as Meta-Llama-3.1-8B-Instruct, gemma-2-9b-it, gemma-2b-it, etc. Each configurations features include dataset name, subset, original dataset index, original dataset split, query ID, completion index, conversation content, completion content, and facts information. The dataset is primarily used for testing and evaluating model performance, particularly in tasks such as dialogue generation, fact extraction, and question answering.