Skip to content

Dataset · FreedomIntelligence

Medical problems with verifiable answers

FreedomIntelligence/medical-o1-verifiable-problem

Open-ended medical questions with reference answers from challenging exams to test LLM reasoning.

What's inside

This dataset contains open-ended medical problems pairing a question with a reference answer sourced from challenging medical exams. Verifiable answers make it possible to validate LLM outputs and improve medical reasoning. Language: English; size: 10K-100K.

What you can do with it

  • SFT for medical LLMs
  • Verifying and evaluating medical reasoning
  • Question-answering benchmarking
Tasks
Question answering, Text generation
Languages
English
Size
10K–100K rows
Downloads this month
2.5K
Likes
125

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("FreedomIntelligence/medical-o1-verifiable-problem")