Skip to content

Dataset · yatin-superintelligence

Synthetic reasoning and web search dataset for small models

yatin-superintelligence/Edge-Agent-Reasoning-WebSearch-260K

A 700+ million token corpus for training compact models in step-by-step task analysis and web search.

What's inside

A synthetically generated English corpus ranging from 100 thousand to 1 million records (over 700 million tokens), including text, images, documents, and 3D data. The materials are designed to teach small language models (SLMs) to decompose complex tasks and recognize a lack of context instead of hallucinating.

What you can do with it

  • Training and SFT of compact local models (SLMs) for device deployment
  • Developing autonomous AI agents with web search and self-verification skills
Tasks
Text generation, Question answering, Multimodal, Robotics
Languages
English
Size
100K–1M rows
Downloads this month
6.8K
Likes
55

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("yatin-superintelligence/Edge-Agent-Reasoning-WebSearch-260K")