Skip to content

Dataset · Locutusque

Large Collection of Instructional Textbooks

Locutusque/UltraTextbooks

A large collection of texts for training language models, including code.

What's inside

UltraTextbooks is a collection of texts for training language models, described as an extensive assembly of high-quality synthetic and human materials. Metadata specifies English and code languages, a size of 1-10 million examples, and text format. Exact composition and source are not disclosed in the provided card fragment.

What you can do with it

  • Language model fine-tuning on instructional textbooks
  • LoRA fine-tuning for text and code
  • Testing model knowledge after training
Tasks
Text generation
Languages
English, Code
Size
1M–10M rows
Downloads this month
1.1K
Likes
209

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("Locutusque/UltraTextbooks")