RuntimeError: (NotFound) Operator (one_hot) is not registered. [Hint: op_info_ptr should not be null.] (at ..\paddle/fluid/framework/op_info.h:152)

[Idaydayup]

代码

import gradio as gr
from paddlenlp import Taskflow
import numpy as np
from PIL import Image
import uuid

初始化文档智能任务模型

docprompt = Taskflow("document_intelligence")

定义模型推理函数

def model_inference(image, prompt):
prompt = prompt.split("?")[:-1]
img = Image.fromarray(np.uint8(image))
filename = "image/" + str(uuid.uuid4()) + ".png"
img.save(filename)
res = docprompt([{"doc": filename, "prompt": prompt}])
json_out = {"result": res}
print(json_out)
return image, json_out

定义清除函数

def clear_all():
return None, None, None

使用Gradio构建用户界面

with gr.Blocks() as demo:
gr.Markdown("ERNIE-Layout")
with gr.Column(scale=1, min_width=100):
# 输入图片和文本框
img_in = gr.Image(value="图片.jpg", label="Input")
text = gr.Textbox(value="发票号码是多少?校验码是多少?", label="输入问题:", lines=2)
with gr.Row():
# 清除和提交按钮
btn1 = gr.Button("Clear")
btn2 = gr.Button("Submit")
json_out = gr.JSON(label="Information Extraction Output")
img_out = gr.Image(label="Output") # 输出图片
# 绑定按钮点击事件到函数
btn1.click(fn=clear_all, inputs=None, outputs=[img_in, img_out, json_out])
btn2.click(fn=model_inference, inputs=[img_in, text], outputs=[img_out, json_out])
demo.launch(server_port=7008) # 启动Gradio界面，监听7008端口)

#报错D:\anaconda\envs\p5\python.exe D:\pycharm\实训2\任务5\内容识别.py
D:\anaconda\envs\p5\lib\site-packages_distutils_hack_init_.py:33: UserWarning: Setuptools is replacing distutils.
warnings.warn("Setuptools is replacing distutils.")
E1228 17:47:01.441207 15760 analysis_config.cc:653] Please compile with MKLDNN first to use MKLDNN
Traceback (most recent call last):
File "D:\pycharm\实训2\任务5\内容识别.py", line 18, in
docprompt = Taskflow("document_intelligence")
File "D:\anaconda\envs\p5\lib\site-packages\paddlenlp\taskflow\taskflow.py", line 804, in init
self.task_instance = task_class(
File "D:\anaconda\envs\p5\lib\site-packages\paddlenlp\taskflow\document_intelligence.py", line 68, in init
self._get_inference_model()
File "D:\anaconda\envs\p5\lib\site-packages\paddlenlp\taskflow\task.py", line 372, in _get_inference_model
self._prepare_static_mode()
File "D:\anaconda\envs\p5\lib\site-packages\paddlenlp\taskflow\task.py", line 227, in _prepare_static_mode
self.predictor = paddle.inference.create_predictor(self._config)
RuntimeError: (NotFound) Operator (one_hot) is not registered.
[Hint: op_info_ptr should not be null.] (at ..\paddle/fluid/framework/op_info.h:152)

进程已结束,退出代码1

[likenamehaojie]

我也遇到了类似的问题。

[m7mdhka]

+1 same issue

[magicianCoder]

应该是paddle版本导致，退回到paddlepaddle==1.5.0就好了（参照taskflow无法使用document_intelligent）

[QQbbla]

回退倒2.5之后会出现别的error，pipe = DocPipeline(preprocessor=preprocessor, docreader=docprompter)
TypeError: init() got an unexpected keyword argument 'docreader'
how to solve it?

[jonyllamado]

Try downgrading paddlepaddle-gpu from 2.6.2 to 2.5.2

python3 -m pip install paddlepaddle-gpu==2.5.2.post120 -f https://www.paddlepaddle.org.cn/whl/linux/mkl/avx/stable.html