EAD/XML Dataset for Prompt-Completion Tasks
Overview
This dataset is specifically designed for text generation and completion tasks involving Encoded Archival Description (EAD) files in XML format. The dataset provides four distinct types of prompt-completion pairs, each crafted to help train and evaluate language models in understanding hierarchical XML structures, element relationships, path-based predictions, and masked element generation.
Dataset summary… See the full description on the dataset page: https://huggingface.co/datasets/Geraldine/Ead-Instruct-full-175k.
