# Multi-source Medical Data

MedAlpaca 13b is a large language model specifically fine-tuned for medical domain tasks, based on the LLaMA architecture with 13 billion parameters, designed to enhance performance in medical Q&A and dialogue tasks.

MedBERT is a Transformer-based pretrained language model specifically designed for biomedical named entity recognition tasks. It is initialized based on Bio_ClinicalBERT and pretrained on multiple biomedical datasets.