Dataset · argilla
21M detailed personas for synthetic data
argilla/FinePersonas-v0.1
A collection of 21 million detailed persona descriptions for controlled synthetic text generation.
What's inside
FinePersonas is an open dataset featuring 21 million detailed personas. It is used to inject unique character traits into text generation systems and yield more diverse synthetic data without manual attribute description. The language is English.
What you can do with it
- Generating synthetic SFT data with specified personas
- Controlled text generation
- Auditing diversity and bias in synthetic data
- Tasks
- Text generation
- Languages
- English
- Size
- 10M–100M rows
- Downloads this month
- 1.2K
- Likes
- 421
Models trained on it
- ringusTheImp/darkScribe-daredevil-8B-abliterated 58
- agency42/Helios-GPT-OSS-20BText generation 24
- Vedant3907/Llama-3.2-1B-PersonaClassifierText generation 16
- Bossmarc747/innovatehub-ai-modelMultimodal 0
- EVA787797/898988Text → image 0
- Reverse-project/project 0
- Cloudpia/AnisQuestion answering 0
- rfnkyz/realistic-test-1Text → image 0
How to load it
from datasets import load_dataset
ds = load_dataset("argilla/FinePersonas-v0.1")