L

Llama3 Med42 8B

Developed by m42-health
Med42-v2 is a clinically aligned large language model suite developed by M42, based on the LLaMA-3 architecture, available in 8B or 70B parameter versions, designed to provide high-quality medical Q&A capabilities.
Downloads 6,755
Release Time : 7/2/2024

Model Overview

An open-access suite of clinical large language models that expands medical knowledge accessibility through instruction and preference fine-tuning, supporting medical Q&A, medical record summarization, and other healthcare-related tasks.

Model Features

Outstanding Clinical Performance
The 70B version achieved a score of 79.10 in the MedQA zero-shot test, surpassing mainstream models like GPT-4.0.
Safety-First Design
Built-in safety mechanisms prevent the generation of harmful medical advice, emphasizing ethics and risk warnings.
Multiple Size Options
Available in 8B and 70B parameter sizes to accommodate different computational resource needs.

Model Capabilities

Medical Q&A
Medical Record Summarization
Diagnostic Assistance
Health Consultation

Use Cases

Clinical Decision Support
Symptom Analysis
Provides possible etiology analysis based on patient-described symptoms.
Achieved an accuracy rate of 73.20% in the MedMCQA test.
Medical Education
Exam Question Solutions
Explains questions from medical exams like USMLE.
Achieved an accuracy rate of 83.80% in the USMLE test.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase