# epfLLM/meditron

Meditron is a suite of open-source medical Large Language Models (LLMs).

Repository: https://github.com/epfLLM/meditron
Canonical: https://ross.abutalabs.com/products/meditron
Homepage: https://huggingface.co/epfl-llm
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2024-04-10T11:48:37+00:00

## Health v2 (maintenance only)
Score: 27/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 72
- inputs: {"age_days": 1014, "days_push": 875, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2208, forks 217 (observed 2026-08-28T04:06:26.148478+00:00)

## What it is
Meditron is a suite of open-source medical large language models (7B and 70B) adapted from Llama-2 via continued pretraining on a curated medical corpus including PubMed content and clinical guidelines. The repository provides code and instructions for downloading and preprocessing the training data and loading the models via Hugging Face Transformers.

## Use cases
- run a medical domain LLM locally
- answer medical exam-style reasoning questions with an open model
- preprocess PubMed abstracts and clinical guidelines for LLM pretraining
- finetune a medical base model on clinical data
- compare open medical LLMs against GPT-3.5 and Flan-PaLM
- download the GAP-Replay medical pretraining corpus

## When to choose
- you need an open-weights LLM with strong medical knowledge for research
- you want a base model to finetune for healthcare NLP tasks
- you need the clinical guidelines corpus or GAP-Replay dataset for your own pretraining

## When to avoid
- you need a production clinical decision-support system - the authors warn against deployment without extensive alignment and trials
- you need up-to-date medical information past the August 2023 knowledge cutoff
- you lack GPU resources for 7B/70B inference

## Facets
- artifact type: library
- maturity: maintenance
- function: llm-training, machine-learning, nlp, etl
- domain: healthcare, large-language-models, deep-learning
- platform: python
- tags: medical-llm, llama-2, continued-pretraining, huggingface, clinical-guidelines, domain-adaptive-pretraining, natural-language-processing, gpu, linux

## Member repositories
- epfLLM/meditron (main) score 27

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:26.148478+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:46:20.474057+00:00, confidence not recorded.
  - readme: https://github.com/epfLLM/meditron (fetched 2026-08-28T04:06:26.148478+00:00, sha 5fa218a69e37)
  - homepage: https://huggingface.co/epfl-llm (fetched 2026-08-29T10:26:45.583198+00:00, sha ab98ff7372fa)
  - site_page: https://huggingface.co/docs (fetched 2026-08-29T10:26:45.592070+00:00, sha bdec26667b98)
  - site_page: https://huggingface.co/docs/hub/organizations-cards (fetched 2026-08-29T10:26:45.596008+00:00, sha 635040b5f704)
  - site_page: https://huggingface.co/pricing (fetched 2026-08-29T10:26:45.594001+00:00, sha de6b7a178be5)
  - site_page: https://huggingface.co/huggingface (fetched 2026-08-29T10:26:45.597631+00:00, sha 0bb9548e5c57)
- Data as of 2026-08-30T08:39:29.467469+00:00.