The DPO dataset is designed to enhance the fairness and ethical alignment of User-VLM models. It is composed of two primary sub-datasets: BiasVision-DPO and VLM-DPO, each contributing unique attributes to improve model performance and reduce biases.
Dataset Details
1. BiasVision-DPO
BiasVision-DPO consists of 12K entries that integrate data from the FairFace and Bias-DPO datasets. This sub-dataset is… See the full description on the dataset page: https://huggingface.co/datasets/ACIDE/user-vlm-dpo.
