AI-in-Health/MedLLMsPracticalGuide

Organizes medical LLM development across three construction paradigms—pre-training from scratch, fine-tuning general models, and prompting—alongside curated clinical datasets and evaluation benchmarks for discriminative and generative tasks. Covers deployment patterns including retrieval-augmented generation, clinical decision support, and automated report generation, while systematically addressing domain-specific challenges like hallucination, knowledge adaptation, and regulatory compliance. Resources span from foundational model architecture guidance to practical multi-modal and agentic approaches for clinical integration.

1,991 stars. No commits in the last 6 months.