README.md

metadata

license: llama3
language:
  - en
  - ja
  - zh
base_model:
  - meta-llama/Meta-Llama-3-8B
pipeline_tag: text-generation
library_name: transformers

Model Card for Model ID

This modelcard aims to be a base template for new models. It has been generated using this raw template.

Model Details

Model Description

• Developed by: [More Information Needed]
• Funded by [optional]: [More Information Needed]
• Shared by [optional]: [More Information Needed]
• Model type: [More Information Needed]
• Language(s) (NLP): [More Information Needed]
• License: [More Information Needed]
• Finetuned from model [optional]: [More Information Needed]

Model Sources [optional]

• Repository: [More Information Needed]
• Paper [optional]: [More Information Needed]
• Demo [optional]: [More Information Needed]

Uses

Direct Use

[More Information Needed]

Downstream Use [optional]

[More Information Needed]

Out-of-Scope Use

[More Information Needed]

Bias, Risks, and Limitations

[More Information Needed]

Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.

How to Get Started with the Model

Use the code below to get started with the model.

[More Information Needed]

Training Details

Training Data

[More Information Needed]

Training Procedure

Preprocessing [optional]

[More Information Needed]

Training Hyperparameters

• Training regime: [More Information Needed]

Speeds, Sizes, Times [optional]

[More Information Needed]

Evaluation

Testing Data, Factors & Metrics

Testing Data

[More Information Needed]

Factors

[More Information Needed]

Metrics

[More Information Needed]

Results

	EN
Model	MedQA-4op	MedQA	MMLU	MedMCQA	PubMedQA
Gemma -7B	52.92	47.56	69.74	48.67	73.44
Llama2 -7B	36.59	28.79	44.19	36.27	32.80
Llama3 -8B	58.52	52.76	70.02	55.25	75.05
Swallow -8B -v0.1	47.87	45.66	65.87	49.99	64.39
med alpaca -7B	37.62	30.99	51.75	34.23	39.84
Meditron -7B	35.09	29.10	46.96	29.88	20.52
Open BioLLM -8B	40.14	50.08	73.15	56.23	65.39
DISC -MedLLM	37.46	32.89	50.00	36.96	33.20
Apollo -7B	54.97	49.68	68.73	53.48	75.86
ELAINE -medLLM	56.15	50.39	67.62	53.74	71.83
ELAINE -medLLM -instruct	58.36	55.84	72.79	54.48	73.24

|

	ZH			JA
Model	MedQA-4op	MedQA	CMExam	JJSIMQA	IgakuQA	DenQA
Gemma -7B	48.90	44.55	38.60	27.61	35.80	25.14
Llama2 -7B	29.52	24.85	23.55	12.61	17.45	17.08
Llama3 -8B	51.42	44.64	39.41	28.91	33.20	23.47
Swallow -8B -v0.1	47.35	40.84	35.94	37.83	45.15	29.03
med alpaca -7B	30.81	25.17	23.40	14.35	16.55	10.83
Meditron -7B	31.10	24.47	22.61	13.48	18.15	15.56
Open BioLLM -8B	50.37	42.59	24.59	20.87	31.50	14.44
DISC -MedLLM	47.18	46.13	41.57	23.26	27.15	23.47
Apollo -7B	65.19	60.98	51.40	25.00	37.40	24.72
ELAINE -medLLM	57.50	52.44	44.99	35.65	45.75	29.86
ELAINE -medLLM -instruct	61.59	55.71	47.19	35.22	46.35	32.36

Summary

Model Examination [optional]

[More Information Needed]

Environmental Impact

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

• Hardware Type: [More Information Needed]
• Hours used: [More Information Needed]
• Cloud Provider: [More Information Needed]
• Compute Region: [More Information Needed]
• Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Glossary [optional]

[More Information Needed]

More Information [optional]

[More Information Needed]

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

import string, time
from vllm import LLM, SamplingParams
import torch

model_name = "kenyano/ELAINE_medLLM"

vllm_paralell = 1

questions_ja = [
    "尿酸値の値はどこまでが正常値ですか？",
]

questions_en = [
    "What is the normal level of uric acid levels?" ,
]

questions_zh = [
    "尿酸的正常水平是多少？",
]


llm = LLM(model=model_name,
            trust_remote_code=True,
            tensor_parallel_size=vllm_paralell,
            dtype="half",
            max_model_len=8192)

sampling_params = SamplingParams(temperature=0.2, top_p=0.8, max_tokens=200, min_tokens=50)

def generate(questions):

    prompts = [f"Human: \n{question}\n\nAssistant: \n" for question in questions]
    outputs = llm.generate(prompts,sampling_params)

    for i, output in enumerate(outputs):
        prompt = output.prompt
        generated_text = output.outputs[0].text

        print("-"*5, "prompt", "-"*5)
        print(f'{prompt}')
        print("-"*5, "generaated", "-"*5)
        print(f'{generated_text}\n')


generate(questions_ja)
generate(questions_en)
generate(questions_zh)

```python