app/model.py

from transformers import AutoModel
from PIL import Image
import requests
from io import BytesIO

# Load model
model = AutoModel.from_pretrained('jinaai/jina-clip-v2', trust_remote_code=True)

def get_text_embedding(texts, truncate_dim=512):
    return model.encode_text(texts, truncate_dim=truncate_dim).tolist()

def get_image_embedding(image_sources, truncate_dim=512):
    processed_images = []
    for src in image_sources:
        if src.startswith('http'):
            response = requests.get(src)
            img = Image.open(BytesIO(response.content)).convert('RGB')
            processed_images.append(img)
    return model.encode_image(processed_images, truncate_dim=truncate_dim).tolist()