PP-DoclayoutV2模型报错

[LinSueyu]

🔎 Search before asking

• [x] I have searched the PaddleOCR Docs and found no similar bug report.
• [x] I have searched the PaddleOCR Issues and found no similar bug report.
• [x] I have searched the PaddleOCR Discussions and found no similar bug report.

🐛 Bug (问题描述)

a100环境中可以下载模型但是运行不了官方样例 拉的官方镜像 ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/paddle:3.0.0-gpu-cuda12.6-cudnn9.5-trt10.5

λ localhost /workspace/code python paddleocr_test_1103.py grep: warning: GREP_OPTIONS is deprecated; please use an alias or script Model files already exist. Using cached files. To redownload, please delete the directory manually: /root/.paddlex/official_models/PP-DocLayoutV2. Traceback (most recent call last): File "/workspace/code/paddleocr_test_1103.py", line 4, in model = LayoutDetection(model_name="PP-DocLayoutV2") File "/usr/local/lib/python3.10/dist-packages/paddleocr/_models/_object_detection.py", line 44, in init super().init(**kwargs) File "/usr/local/lib/python3.10/dist-packages/paddleocr/_models/base.py", line 46, in init self.paddlex_predictor = self._create_paddlex_predictor() File "/usr/local/lib/python3.10/dist-packages/paddleocr/_models/base.py", line 76, in _create_paddlex_predictor return create_predictor( File "/usr/local/lib/python3.10/dist-packages/paddlex/inference/models/init.py", line 84, in create_predictor return BasePredictor.get(model_name)( File "/usr/local/lib/python3.10/dist-packages/paddlex/inference/models/object_detection/predictor.py", line 112, in init self.pre_ops, self.infer, self.post_op = self._build() File "/usr/local/lib/python3.10/dist-packages/paddlex/inference/models/object_detection/predictor.py", line 143, in _build infer = self.create_static_infer() File "/usr/local/lib/python3.10/dist-packages/paddlex/inference/models/base/predictor/base_predictor.py", line 302, in create_static_infer self.model_name, self.model_dir, self.MODEL_FILE_PREFIX, self._pp_option AttributeError: 'DetPredictor' object has no attribute '_pp_option'. Did you mean: 'pp_option'?

🏃‍♂️ Environment (运行环境)

ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/paddle:3.0.0-gpu-cuda12.6-cudnn9.5-trt10.5 PaddlePaddle 3.0.0 PaddleX 3.3.6 paddleocr 3.3.1

🌰 Minimal Reproducible Example (最小可复现问题的Demo)

from paddleocr import LayoutDetection

model = LayoutDetection(model_name="PP-DocLayoutV2")

output = model.predict("./demo.jpg", batch_size=1, layout_nms=True) for res in output: res.print() res.save_to_img(save_path="/workspace/") res.save_to_json(save_path="/workspace/paddleocr_test.json")

[changdazhou]

建议您更新paddlex到最新3.3.8版本后重试哈，该问题已修复

[DooFy2020]

建议您更新paddlex到最新3.3.8版本后重试哈，该问题已修复

我遇到了类似的问题，使用的是PaddleORC VL。 paddelx 3.3.8 paddleocr 3.3.1 paddlepaddle-gpu 3.2.1

https://github.com/PaddlePaddle/PaddleOCR/issues/16936

[changdazhou]

感谢您的反馈，我们会尽快排查这个问题

[Bobholamovic]

建议您更新paddlex到最新3.3.8版本后重试哈，该问题已修复

我遇到了类似的问题，使用的是PaddleORC VL。 paddelx 3.3.8 paddleocr 3.3.1 paddlepaddle-gpu 3.2.1

#16936

请问是否可以按照如下步骤确认下代码中的关键路径是否被执行到：

1. pip show paddlex确定PaddleX源代码位置（Location），比如/usr/local/lib/python3.10/dist-packages/paddlex。

2. 找到paddlex/inference/models/base/predictor/base_predictor.py，在第161行添加断点：

               if not use_hpip:
                   breakpoint()
                   self._pp_option = self._prepare_pp_option(pp_option, device)
   

   如果第161行不是self._pp_option = self._prepare_pp_option(pp_option, device)，说明使用的PaddleX版本很可能并不是3.3.8，需要排查本地是否存在多个版本的PaddleX。

3. 观察断点位置是否被执行到。

[TingquanGao]

The issue has no response for a long time and will be closed. You can reopen or new another issue if are still confused.

---

From Bot