app.py

import gradio as gr
import argparse
import torch
from PIL import Image
from donut import DonutModel
def demo_process(input_img):
    global model, task_prompt, task_name
    input_img = Image.fromarray(input_img)
    output = model.inference(image=input_img, prompt=task_prompt)["predictions"][0]
    return output
parser = argparse.ArgumentParser()
parser.add_argument("--task", type=str, default="Booking")
parser.add_argument("--pretrained_path", type=str, default="result/train_booking/20241112_150925")
args, left_argv = parser.parse_known_args()
task_name = args.task
task_prompt = f"<s_{task_name}>"
model = DonutModel.from_pretrained("./result/train_booking/20241112_150925")
if torch.cuda.is_available():
    model.half()
    device = torch.device("cuda")
    model.to(device)
else:
    model.encoder.to(torch.bfloat16)
model.eval()
demo = gr.Interface(fn=demo_process,inputs="image",outputs="json", title=f"Donut 🍩 demonstration for `{task_name}` task",)
demo.launch(debug=True)