VLFeedback

Dataset Card for VLFeedback Homepage: Repository: Paper: Dataset Summary VLFeedback is a large-scale vision-language preference dataset, annotated by GPT-4V. It consists of 80k multi-modal instructions from various souces that encompass various capabilities of LVLMs. We build a model pool of 12 LVLMs and each data sample contains 4 responses from different… See the full description on the dataset page:

Source: Hugging Face Hub (MMInstruction/VLFeedback). Metadata imported from the dataset’s Hub tags.