app.py

import gradio as gr
import torch
from PIL import Image

from donut import DonutModel

def demo_process(input_img):
    global pretrained_model, task_prompt, task_name
    # input_img = Image.fromarray(input_img)
    output = pretrained_model.inference(image=input_img, prompt=task_prompt)["predictions"][0]
    return output

task_prompt = f"<s_cord-v2>"

image = Image.open("./sample_1.jpg")
image.save("cord_sample_1.png")
image = Image.open("./sample_2.jpg")
image.save("cord_sample_2.png")

pretrained_model = DonutModel.from_pretrained("thinkersloop/donut-demo")
pretrained_model.encoder.to(torch.bfloat16)
pretrained_model.eval()

demo = gr.Interface(
    fn=demo_process,
    inputs= gr.inputs.Image(type="pil"),
    outputs="json",
    title=f"Transformers demo for `cord-v2` task",
    description="""This model is trained with 30 driver's license images of CORD dataset. <br>""",
    examples=[["cord_sample_1.png"], ["cord_sample_2.png"]],
    cache_examples=False,
)

demo.launch()