chore: upgrade to MedGemma-4B-IT for medical analysis

- Add HF_TOKEN authentication for gated model access
- Use Google MedGemma-4B-IT (medical specialist model)
- Add huggingface_hub dependency for authentication
- Instructions for setting HF_TOKEN in Spaces secrets

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

app.py

@@ -2,6 +2,7 @@ import json
 import re
 from typing import List, Optional, Tuple
 import numpy as np
+import os
 
 import gradio as gr
 import spaces
@@ -9,6 +10,12 @@ import torch
 from PIL import Image
 from transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessor, AutoTokenizer, AutoModelForCausalLM
 from qwen_vl_utils import process_vision_info
+from huggingface_hub import login
+
+# Hugging Face 토큰으로 로그인 (Spaces Secret에서 가져옴)
+HF_TOKEN = os.getenv("HF_TOKEN")
+if HF_TOKEN:
+    login(token=HF_TOKEN)
 
 # OCR 모델 ID
 OCR_MODEL_ID = "Qwen/Qwen2.5-VL-7B-Instruct"
@@ -371,6 +378,9 @@ with gr.Blocks(theme=gr.themes.Soft(), css=CUSTOM_CSS) as demo:
     **🤖 기술 스택**
     - Qwen2.5-VL-7B-Instruct (OCR 텍스트 추출)
     - Google MedGemma-4B-IT (의료 전문 모델 - 약 정보 분석 및 설명)
+
+    **🔑 설정 방법**
+    - Hugging Face Spaces의 Settings → Repository secrets에서 `HF_TOKEN` 추가 필요
     """)
 
 if __name__ == "__main__":

requirements.txt

@@ -6,3 +6,4 @@ Pillow
 numpy
 qwen-vl-utils
 accelerate
+huggingface_hub