There are three types of datasets, namely ASQA, QAMPARI, and ELI5. We provide their raw data and the data with summary and answer generated by the model.
raw data: We put the raw data in the origin directory. You can also find them and get more information in the repo of ALCE.
summary-answer data: We put the data with summary and answer generated by the model(gpt-3.5-turbo-0301)in the summary-answer directory.… See the full description on the dataset page:
https://huggingface.co/datasets/BeastyZ/LLM-Verified-Retrieval.
