README.md

metadata

annotations_creators:
  - expert-generated
language:
  - en
license: mit
multilinguality:
  - monolingual
size_categories:
  - 1K<n<10K
source_datasets:
  - original
task_categories:
  - question-answering
task_ids:
  - closed-domain-qa
pretty_name: Eye Disease QA Dataset

Eye Disease QA Dataset

This dataset is designed for training and evaluating large language models (LLMs) in the field of ophthalmology. It consists of a structured disease knowledge base and question-answer (QA) pairs derived from that knowledge. The dataset can be used for supervised fine-tuning, testing, and knowledge-enhanced retrieval tasks.

📂 Files Included

• eye_disease_knowledge_base.json A curated knowledge base covering common eye diseases such as glaucoma, cataracts, and retinal disorders. Entries are extracted from clinical guidelines, textbooks, and authoritative online sources. This file can serve as the grounding corpus for retrieval-augmented generation (RAG) models.
• eye_QA_train.json A collection of question-answer pairs constructed based on the knowledge base. Designed for fine-tuning LLMs. Each entry includes a question, a corresponding answer, and optionally the source or evidence.
• eye_QA_test.json A held-out evaluation set of 500 QA pairs to test the generalization ability of QA models. The format is consistent with the training set.

✅ Use Cases

• Fine-tuning instruction-tuned language models for medical QA
• Building retrieval-augmented generation (RAG) systems
• Evaluating knowledge-grounded medical dialogue agents

📌 License

This dataset is provided for research purposes only. Please verify medical content before any clinical use.

📞 Contact

For questions or collaborations, please contact the dataset maintainer via your Hugging Face profile or email.