app.py

from sentence_transformers import SentenceTransformer
import gradio as gr

# Load once
image_model = SentenceTransformer("clip-ViT-B-32")
text_model = SentenceTransformer("clip-ViT-B-32-multilingual-v1")

def get_image_embedding(image):
    emb = image_model.encode(image)
    return {"embedding": emb.tolist()}

def get_text_embedding(text):
    emb = text_model.encode(text)
    return {"embedding": emb.tolist()}

with gr.Blocks() as demo:
    with gr.Tab("Image Embedding"):
        img_input = gr.Image(type="pil")
        img_output = gr.JSON()
        img_btn = gr.Button("Generate")
        img_btn.click(get_image_embedding, img_input, img_output)

    with gr.Tab("Text Embedding"):
        text_input = gr.Textbox()
        text_output = gr.JSON()
        text_btn = gr.Button("Generate")
        text_btn.click(get_text_embedding, text_input, text_output)

demo.launch()