HectorRguez/wildchat-advertisement-placement-preference

Advertisement Integration Preference Dataset 1. Overview This dataset contains approximately 3,000 preference samples designed to train models on appropriate advertisement integration in conversational responses. Each sample compares two versions of a response, and it is annotated as follows: response_1: The first response is better response_2: The second response is better both_good: Both responses meet the evaluation criteria both_bad: Both responses fail to… See the full description on the dataset page: https://huggingface.co/datasets/HectorRguez/wildchat-advertisement-placement-preference.