Skip to content

Dataset · RodrigoFlorencio

Portuguese health Q&A corpus

RodrigoFlorencio/duvidas-saude-ptbr

Over 33,000 real patient questions and 127,000 doctor answers in Portuguese.

What's inside

The corpus contains over 33,000 unique real patient questions across 62 pathologies, collected from public Brazilian web sites, alongside over 127,000 medical professional answers. The text dataset covers Q&A and text generation in Portuguese, and includes artifacts for SFT and LoRA models.

What you can do with it

  • Fine-tuning medical language models (SFT/LoRA) in Portuguese
  • Developing medical query routing systems by specialization
Tasks
Question answering, Text generation
Languages
Portuguese
Size
100K–1M rows
Downloads this month
2.7K
Likes
2

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("RodrigoFlorencio/duvidas-saude-ptbr")