Skip to content

Dataset · sedthh

English books from Project Gutenberg

sedthh/gutenberg_english

Collection of 48,284 English books from Project Gutenberg, prepared for generative model training.

What's inside

The dataset contains 48,284 rows of English e-book texts from Project Gutenberg (over 80% of all books in this language from the library). Original text metadata was extracted and placed into a separate METADATA column as serialized JSON. The corpus was initially collected for the OpenAssistant project.

What you can do with it

  • Pre-training and fine-tuning language models on long cohesive texts
  • Training fiction generation models and stylistic LoRAs
Tasks
Text generation
Languages
English
Size
10K–100K rows
Downloads this month
35.4K
Likes
41

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("sedthh/gutenberg_english")