Skip to content

Dataset · Manusagents

Distillation dataset from 73 models

Manusagents/GPT-5.5-Gemini-3.1-Pro-Grok-4-Claude-Fable-5-Mythos-5-Qwen-3.7-Max-and-more-Distillation-Dataset

18M+ distillation signals from 73 open models and 7k GitHub repositories for training text models.

What's inside

Text distillation corpus: 18M+ signals collected from 7,090 source GitHub repositories across 8 categories, totaling ~76GB+. According to the card, the data combines outputs from 73 open models in English and multilingual formats. It is an assembled training dataset rather than human-written text.

What you can do with it

  • SFT for text LLMs using distillation signals
  • LoRA fine-tuning for code and reasoning
  • Testing distillation pipelines
Tasks
Text generation, Other
Languages
English, Multilingual
Size
10M–100M rows
Downloads this month
18.2K
Likes
290

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("Manusagents/GPT-5.5-Gemini-3.1-Pro-Grok-4-Claude-Fable-5-Mythos-5-Qwen-3.7-Max-and-more-Distillation-Dataset")