84 lines
5.3 KiB
Markdown
84 lines
5.3 KiB
Markdown
---
|
||
license: llama3
|
||
datasets:
|
||
- Henrychur/MMedC
|
||
language:
|
||
- en
|
||
- zh
|
||
- ja
|
||
- fr
|
||
- ru
|
||
- es
|
||
tags:
|
||
- medical
|
||
---
|
||
# MMedLM
|
||
[💻Github Repo](https://github.com/MAGIC-AI4Med/MMedLM) [🖨️arXiv Paper](https://arxiv.org/abs/2402.13963)
|
||
|
||
The official model weights for "Towards Building Multilingual Language Model for Medicine".
|
||
|
||
|
||
## Introduction
|
||
This repo contains MMed-Llama 3, a multilingual medical foundation model with 8 billion parameters. MMed-Llama 3 builds upon the foundation of Llama 3 and has been further pretrained on MMedC, a comprehensive multilingual medical corpus. This further pretraining enhances the model's medical-domain knowledge.
|
||
|
||
The model underwent further pretraining on MMedC with the following hyperparameters:
|
||
- Iterations: 15000
|
||
- Global batch size: 512
|
||
- Cutoff length: 8192
|
||
- Learning rate: 2e-5
|
||
|
||
The model can be loaded as follows:
|
||
```py
|
||
import torch
|
||
from transformers import AutoTokenizer, AutoModelForCausalLM
|
||
tokenizer = AutoTokenizer.from_pretrained("Henrychur/MMed-Llama-3-8B")
|
||
model = AutoModelForCausalLM.from_pretrained("Henrychur/MMed-Llama-3-8B", torch_dtype=torch.float16)
|
||
```
|
||
|
||
- Note that this is a foundation model that has not undergone instruction fine-tuning.
|
||
## News
|
||
[2024.2.21] Our pre-print paper is released ArXiv. Dive into our findings [here](https://arxiv.org/abs/2402.13963).
|
||
|
||
[2024.2.20] We release [MMedLM](https://huggingface.co/Henrychur/MMedLM) and [MMedLM 2](https://huggingface.co/Henrychur/MMedLM2). With an auto-regressive continues training on MMedC, these models achieves superior performance compared to all other open-source models, even rivaling GPT-4 on MMedBench.
|
||
|
||
[2023.2.20] We release [MMedC](https://huggingface.co/datasets/Henrychur/MMedC), a multilingual medical corpus containing 25.5B tokens.
|
||
|
||
[2023.2.20] We release [MMedBench](https://huggingface.co/datasets/Henrychur/MMedBench), a new multilingual medical multi-choice question-answering
|
||
benchmark with rationale. Check out the leaderboard [here](https://henrychur.github.io/MultilingualMedQA/).
|
||
|
||
## Evaluation on MMedBench
|
||
The further pretrained MMedLM 2 showcast it's great performance in medical domain across different language.
|
||
|
||
| Method | Size | Year | MMedC | MMedBench | English | Chinese | Japanese | French | Russian | Spanish | Avg. |
|
||
|------------------|------|---------|-----------|-----------|----------------|----------------|----------------|----------------|----------------|----------------|----------------|
|
||
| GPT-3.5 | - | 2022.12 | ✗ | ✗ | 56.88 | 52.29 | 34.63 | 32.48 | 66.36 | 66.06 | 51.47 |
|
||
| GPT-4 | - | 2023.3 | ✗ | ✗ | 78.00 | 75.07 | 72.91 | 56.59 | 83.62 | 85.67 | 74.27 |
|
||
| Gemini-1.0 pro | - | 2024.1 | ✗ | ✗ | 53.73 | 60.19 | 44.22 | 29.90 | 73.44 | 69.69 | 55.20 |
|
||
| BLOOMZ | 7B | 2023.5 | ✗ | trainset | 43.28 | 58.06 | 32.66 | 26.37 | 62.89 | 47.34 | 45.10 |
|
||
| InternLM | 7B | 2023.7 | ✗ | trainset | 44.07 | 64.62 | 37.19 | 24.92 | 58.20 | 44.97 | 45.67 |
|
||
| Llama 2 | 7B | 2023.7 | ✗ | trainset | 43.36 | 50.29 | 25.13 | 20.90 | 66.80 | 47.10 | 42.26 |
|
||
| MedAlpaca | 7B | 2023.3 | ✗ | trainset | 46.74 | 44.80 | 29.64 | 21.06 | 59.38 | 45.00 | 41.11 |
|
||
| ChatDoctor | 7B | 2023.4 | ✗ | trainset | 43.52 | 43.26 | 25.63 | 18.81 | 62.50 | 43.44 | 39.53 |
|
||
| PMC-LLaMA | 7B | 2023.4 | ✗ | trainset | 47.53 | 42.44 | 24.12 | 20.74 | 62.11 | 43.29 | 40.04 |
|
||
| Mistral | 7B | 2023.10 | ✗ | trainset | 61.74 | 71.10 | 44.72 | 48.71 | 74.22 | 63.86 | 60.73 |
|
||
| InternLM 2 | 7B | 2024.2 | ✗ | trainset | 57.27 | 77.55 | 47.74 | 41.00 | 68.36 | 59.59 | 58.59 |
|
||
| MMedLM(Ours) | 7B | - | ✓ | trainset | 49.88 | 70.49 | 46.23 | 36.66 | 72.27 | 54.52 | 55.01 |
|
||
| MMedLM 2(Ours) | 7B | - | ✓ | trainset | 61.74 | 80.01 | 61.81 | 52.09 | 80.47 | 67.65 | 67.30 |
|
||
|MMed-Llama 3(Ours)|8B |- | ✓ | trainset | 66.06| 79.25 | 61.81 | 55.63 | 75.39 | 68.38 | 67.75 |
|
||
- GPT and Gemini is evluated under zero-shot setting through API
|
||
- Open-source models first undergo training on the trainset of MMedBench before evaluate.
|
||
|
||
## Contact
|
||
If you have any question, please feel free to contact qiupengcheng@pjlab.org.cn.
|
||
|
||
## Citation
|
||
```
|
||
@misc{qiu2024building,
|
||
title={Towards Building Multilingual Language Model for Medicine},
|
||
author={Pengcheng Qiu and Chaoyi Wu and Xiaoman Zhang and Weixiong Lin and Haicheng Wang and Ya Zhang and Yanfeng Wang and Weidi Xie},
|
||
year={2024},
|
||
eprint={2402.13963},
|
||
archivePrefix={arXiv},
|
||
primaryClass={cs.CL}
|
||
}
|
||
``` |