app.py

import gradio as gr
from gradio_client import Client, handle_file


client_idefics2 = Client("HuggingFaceH4/idefics2-8b-playground")
client_idefics2_dpoed = Client("HuggingFaceH4/idefics2-8b-vdpoed-playground")


def respond(multimodal_input):
    x = {"text": multimodal_input["text"], "files": [handle_file(file) for file in multimodal_input["files"]]}
    text_1 = client_idefics2.predict(x, api_name="/predict")
    text_2 = client_idefics2_dpoed.predict(x, api_name="/predict")
    return text_1, text_2


gr.Interface(
    respond,
    inputs=[gr.MultimodalTextbox(file_types=["image"], show_label=False)],
    outputs=[gr.Textbox(label="idefics2-8b"), gr.Textbox(label="idefics2-8b-dpoed")],
).launch()