A French medical instruction-tuning dataset (~508K examples) compiled from three public medical QA / dialogue sources:
ruslanmv/ai-medical-chatbot — 256,010 examples (config ai_medical_chatbot, default)
lavita/medical-qa-datasets (all-processed config) — 230,041 examples (config medical_qa_datasets)
FreedomIntelligence/Medical-R1-Distill-Data — 21,641 examples (config medical_r1_distill_data)
Each source question was machine-translated into French, then a… See the full description on the dataset page:
https://huggingface.co/datasets/bofenghuang/medical-qa-fr-v0.1.
