[Question]: paddlenlp 3.0.0b1版本使用uie-m-base报错找不到static/inference.pdmodel

[summeryqliu]

请提出你的问题

CPU版本运行官方示例：

from pprint import pprint
from paddlenlp import Taskflow
schema = schema = ['Time', 'Player', 'Competition', 'Score']# Define the schema for entity extraction
ie = Taskflow('information_extraction', schema=schema, model="uie-m-base", schema_lang="en")
pprint(ie("2月8日上午北京冬奥会自由式滑雪女子大跳台决赛中中国选手谷爱凌以188.25分获得金牌！")) # B

报错提示：

Traceback (most recent call last):
  File "/home/yqliu/liuyu_qing/work/study/paddle/PaddleNLP-3.0.0-beta2/legacy/model_zoo/uie/a.py", line 4, i
    ie = Taskflow('information_extraction', schema=schema, model="uie-m-base", schema_lang="en")
  File "/home/yqliu/anaconda3/envs/uie/lib/python3.9/site-packages/paddlenlp/taskflow/taskflow.py", line 809
    self.task_instance = task_class(
  File "/home/yqliu/anaconda3/envs/uie/lib/python3.9/site-packages/paddlenlp/taskflow/information_extraction
    self._get_inference_model()
  File "/home/yqliu/anaconda3/envs/uie/lib/python3.9/site-packages/paddlenlp/taskflow/task.py", line 372, in
    self._prepare_static_mode()
  File "/home/yqliu/anaconda3/envs/uie/lib/python3.9/site-packages/paddlenlp/taskflow/task.py", line 227, in
    self.predictor = paddle.inference.create_predictor(self._config)
RuntimeError: (NotFound) Cannot open file /home/yqliu/.paddlenlp/taskflow/information_extraction/uie-m-base/hether the file is normal.

使用uie-base、uie-base-en模型无问题，混合模型uie-m-base报错，基于uie-m-base微调了一个新模型，微调模型可以执行评估代码eval.py，但是使用taskflow也报相同的错误。

环境信息：

# Name                    Version                   Build  Channel
paddle2onnx               1.3.1                    pypi_0    pypi
paddlefsl                 1.1.0                    pypi_0    pypi
paddlenlp                 3.0.0b1                  pypi_0    pypi
paddlepaddle              3.0.0b1         py39_cpu_many_linux    paddle

[fresh-meet]

Traceback (most recent call last):
  File "/home/XXX/code/PaddleNLP-develop/slm/model_zoo/uie/uie_worlflow.py", line 12, in <module>
    ie = Taskflow('information_extraction', schema=schema, model='uie-base',task_path='./checkpoint/model_best')
  File "/home/XXX/miniconda3/envs/paddle/lib/python3.9/site-packages/paddlenlp/taskflow/taskflow.py", line 809, in __init__
    self.task_instance = task_class(
  File "/home/XXX/miniconda3/envs/paddle/lib/python3.9/site-packages/paddlenlp/taskflow/information_extraction.py", line 536, in __init__
    self._get_inference_model()
  File "/home/XXX/miniconda3/envs/paddle/lib/python3.9/site-packages/paddlenlp/taskflow/task.py", line 371, in _get_inference_model
    self._config = paddle.inference.Config(self._static_model_file, self._static_params_file)
RuntimeError: (NotFound) Cannot open file ./checkpoint/model_best/static/inference.pdmodel, please confirm whether the file is normal.
  [Hint: Expected paddle::inference::IsFileExists(prog_file_) == true, but received paddle::inference::IsFileExists(prog_file_):0 != true:1.] (at ../paddle/fluid/inference/api/analysis_config.cc:111)

========================
使用finetune脚本训练的新模型同样有这样的问题，似乎finetune使用的模型和workflow的模型不一致，请教一下是不是有冲突，怎么解决？

[Fantasy-02]

先设置环境变量
export FLAGS_enable_pir_api=0
再进行微调试试