datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
multidomain-complex-text-pool
Complex Text Pool Dataset
Overview
A curated collection of complex, long-form English texts sampled from 9 diverse domains. Each document has been truncated to a maximum of 4,000 characters, preserving clean sentence boundaries. The dataset is designed to provide challenging, real-world text samples across multiple subject areas.
Categories and Sample Counts
Category
Samples
news
9,999
encyclopedic
10,000
conversational
10,000… See the full description on the dataset page: https://huggingface.co/datasets/Pankaj8922/multidomain-complex-text-pool.synthetic-complex-Text-to-SQL
Synthetic Complex Text-to-SQL
Synthetic Text-to-SQL using multiple joins, WHERE statements, window and aggregate functions on filtered SQL from bigcode/the-stack.
Thesis_Development_of_a_Complex_of_Neural_Networks_for_Linked_Generation_of_Large_TextsHere presented a partially synthesized dataset, developed utilizing the GPT-4 model, for the purpose of NLG, particulary for the task of hierarchical generation of longer texts from short summaries. The creation of this dataset was undertaken as a component of my thesis paper. It incorporates excerpts from prominent British and American novels, from which plots, summaries, and metadata have been derived using GPT-4 API to facilitate extensive future research.
The metadata included in the… See the full description on the dataset page: https://huggingface.co/datasets/Fleur-roar/Thesis_Development_of_a_Complex_of_Neural_Networks_for_Linked_Generation_of_Large_Texts.