[Question]: 使用uie-m-large模型报错

[MachineSheep]

请提出你的问题

完整的报错信息如下

RuntimeError Traceback (most recent call last) Input In [7], in <cell line: 2>() 1 schema = ['Time', "date"] ----> 2 ie2 = Taskflow('information_extraction', schema=schema, model="uie-m-large", schema_lang="en")

File ~/work/202211/event/paddlenlp/taskflow/taskflow.py:543, in Taskflow.init(self, task, model, mode, device_id, **kwargs) 541 self.kwargs = kwargs 542 task_class = TASKS[self.task][tag][self.model]['task_class'] --> 543 self.task_instance = task_class(model=self.model, 544 task=self.task, 545 priority_path=self.priority_path, 546 **self.kwargs) 547 task_list = TASKS.keys() 548 Taskflow.task_list = task_list

File ~/work/202211/event/paddlenlp/taskflow/information_extraction.py:360, in UIETask.init(self, task, model, schema, schema_lang, **kwargs) 358 self._check_task_files() 359 self._check_predictor_type() --> 360 self._get_inference_model() 361 self._usage = usage 362 self._is_en = True if model in ['uie-base-en' 363 ] or schema_lang == 'en' else False

File ~/work/202211/event/paddlenlp/taskflow/task.py:254, in Task._get_inference_model(self) 251 if self._predictor_type == "paddle-inference": 252 self._config = paddle.inference.Config(self._static_model_file, 253 self._static_params_file) --> 254 self._prepare_static_mode() 255 else: 256 self._prepare_onnx_mode()

File ~/work/202211/event/paddlenlp/taskflow/task.py:174, in Task._prepare_static_mode(self) 172 if self.task in ["document_intelligence", "knowledge_mining"]: 173 self._config.switch_ir_optim(False) --> 174 self.predictor = paddle.inference.create_predictor(self._config) 175 self.input_names = [name for name in self.predictor.get_input_names()] 176 self.input_handles = [ 177 self.predictor.get_input_handle(name) 178 for name in self.predictor.get_input_names() 179 ]

RuntimeError: (Unavailable) An error occurred while loading model parameters. Please check whether the model file is complete or damaged. [Hint: Expected static_cast(*buffer) == true, but received static_cast(*buffer):0 != true:1.] (at /paddle/paddle/fluid/operators/load_combine_op.h:86) [operator < load_combine > error]

[linjieccc]

看报错应该是模型文件下载的问题，删除~/.paddlenlp/taskflow/information_extraction重新下载试试

[MachineSheep]

看报错应该是模型文件下载的问题，删除~/.paddlenlp/taskflow/information_extraction重新下载试试

删除了这个文件夹下的uie-m-large文件夹，然后重新下载，还是报错了

[MachineSheep]

看报错应该是模型文件下载的问题，删除~/.paddlenlp/taskflow/information_extraction重新下载试试 uie-m-large文件夹内容如下

[linjieccc]

large模型比较大，有可能还是下载的问题，或者试试手动下载模型文件到这个目录~/.paddlenlp/taskflow/information_extraction/uie-m-large，链接：https://bj.bcebos.com/paddlenlp/taskflow/information_extraction/uie_m_large_v1.0/model_state.pdparams

[MachineSheep]

手动下载之后，windows和linux系统上都还是报最初的错误

[MachineSheep]

large模型比较大，有可能还是下载的问题，或者试试手动下载模型文件到这个目录~/.paddlenlp/taskflow/information_extraction/uie-m-large，链接：https://bj.bcebos.com/paddlenlp/taskflow/information_extraction/uie_m_large_v1.0/model_state.pdparams

model_state.pdparams详细属性如下

[linjieccc]

paddlenlp的版本是多少？

[MachineSheep]

paddlenlp的版本是多少？

2.4.0

[linjieccc]

可以尝试看下model_state.pdparams这个文件对应的md5值是否是 75f3989c515f05f6842e314d3f75ee27

[MachineSheep]

可以尝试看下model_state.pdparams这个文件对应的md5值是否是 75f3989c515f05f6842e314d3f75ee27

md5值没有问题

[myboyliu]

我的也是同样问题。看看解决了吗？

File "D:\Anaconda3\envs\py3.8\lib\site-packages\paddlenlp\taskflow\task.py", line 174, in _prepare_static_mode self.predictor = paddle.inference.create_predictor(self._config) RuntimeError: (Unavailable) An error occurred while loading model parameters. Please check whether the model file is complete or damaged. [Hint: Expected static_cast(*buffer) == true, but received static_cast(*buffer):0 != true:1.] (at ..\paddle/fluid/operators/load_combine_op.h:86) [operator < load_combine > error] python-BaseException

[MachineSheep]

我的也是同样问题。看看解决了吗？

File "D:\Anaconda3\envs\py3.8\lib\site-packages\paddlenlp\taskflow\task.py", line 174, in _prepare_static_mode self.predictor = paddle.inference.create_predictor(self._config) RuntimeError: (Unavailable) An error occurred while loading model parameters. Please check whether the model file is complete or damaged. [Hint: Expected static_cast(*buffer) == true, but received static_cast(*buffer):0 != true:1.] (at ..\paddle/fluid/operators/load_combine_op.h:86) [operator < load_combine > error] python-BaseException

没有解决

[linjieccc]

Hi，可以check下机器的内存使用情况，看下是否有内存不足的问题

[github-actions[bot]]

This issue is stale because it has been open for 60 days with no activity. 当前issue 60天内无活动，被标记为stale。

[github-actions[bot]]

This issue was closed because it has been inactive for 14 days since being marked as stale. 当前issue 被标记为stale已有14天，即将关闭。

[xiahexiang2]

Hi，可以check下机器的内存使用情况，看下是否有内存不足的问题

你好，我也遇到该问题了。请问是内存不足还是显存不足导致？至少需要多少内存或显存才能加载uie-m-larege

[jianxiaojin]

现在解决了吗？