Dataset · AdaptLLM
Medical benchmark for evaluating fine-tuned LLMs
AdaptLLM/medicine-tasks
Collection of medical classification and question-answering tasks from an ICLR 2024 paper to test model knowledge.
What's inside
Contains 1 to 10 thousand examples in English in text and tabular formats. The dataset was prepared by the authors of the ICLR 2024 domain adaptation study for language models and is intended to evaluate their ability to answer questions and classify texts.
What you can do with it
- Evaluating the quality of medical LLMs in question-answering tasks
- Testing models on zero-shot classification of medical texts
- Verifying the retention of instruction-following skills after domain fine-tuning
- Tasks
- Text classification, Question answering, Zero shot classification
- Languages
- English
- Size
- 1K–10K rows
- Downloads this month
- 381
- Likes
- 33
How to load it
from datasets import load_dataset
ds = load_dataset("AdaptLLM/medicine-tasks")