This dataset contains a stratified sample of 60 input texts (prompts) for continuation and their respective original references, selected to cover a variety of linguistic registers and domains. It is designed to be executed within the Galician Rapid Human Evaluation Framework (galician-rapid-human-gen-eval) for LLMs in Galician.
Its primary purpose is to act as an input for the evaluation of model's open-ended… See the full description on the dataset page: https://huggingface.co/datasets/proxectonos/Galician-Generative-Eval-Prompts.
