Note Dataset of patient case reports from PubMed Central.

Note Dataset of clinical guidelines from different international sources.

Note Dataset of multilingual (English, Spanish, Hindi, Chinese) health queries and answers from patients.

Note USMLE Self-Assessments for Steps 1, 2, and 3 (released in 2022).

Note Dataset of synthetically-generated clinical notes.

Note Dataset of synthetically-generated doctor-patient dialogue.

meta-llama/Llama-2-70b-chat-hf

Text Generation • Updated Apr 17 • 204k • 2.16k

Note Highest performing general chat-optimized open-access LLM.

epfl-llm/meditron-70b

Text Generation • Updated Dec 7, 2023 • 474 • 228

Note Highest performing domain-adapted open-access LLM (not chat-optimized).

emilyalsentzer/Bio_ClinicalBERT

Fill-Mask • Updated Mar 31, 2023 • 3.03M • 287

Note Encoder for clinical notes.

ncbi/MedCPT-Query-Encoder

Feature Extraction • Updated Dec 3, 2023 • 348k • 28

Note Encoder for PubMed queries by the NCBI. They also have an article encoder.

microsoft/llava-med-7b-delta

Text Generation • Updated Nov 10, 2023 • 203 • 65

Note Vision-language model domain-adapted for medicine.

HuggingFaceH4/zephyr-7b-beta

Text Generation • Updated Oct 16 • 543k • • 1.61k

Note Strong general 7B chat-optimized LLM.

Note Demo for medical note de-identification algorithm.

Note Demo for comparing outputs of generative biomedical LMs.

Note Larger curated list of clinical and biomedical LMs.