sean0042/KorMedMCQA
Viewer • Updated • 7.49k • 1.71k • 43
How to use ghddls7799/gemma-2-2b-med-ko-qlora with PEFT:
from peft import PeftModel
from transformers import AutoModelForCausalLM
base_model = AutoModelForCausalLM.from_pretrained("google/gemma-2-2b-it")
model = PeftModel.from_pretrained(base_model, "ghddls7799/gemma-2-2b-med-ko-qlora")A Korean medical-domain QLoRA adapter based on
google/gemma-2-2b-it.
ChuGyouk/GenMedGPT-5k-kosean0042/KorMedMCQAKorMedMCQA training used only the train split.
The dev split was used for validation, while test and fewshot
were excluded from training.
This repository contains a PEFT/LoRA adapter, not the complete base
model. Users must separately obtain access to google/gemma-2-2b-it
and agree to the Gemma Terms of Use.
Because KorMedMCQA is marked CC BY-NC 2.0, commercial use and public redistribution require a separate license review.
This model is intended for research and educational evaluation. It is not a medical device and must not be used as a substitute for diagnosis, treatment, or professional medical advice.