PIR 优化尚不支持 Kunlun P800 服务器环境。

Search before asking

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

Bug (Problem description)

Running ppocr-v4 and pp-doclayoutv2: When using a json format model weight, it is not possible to avoid PIR optimization, which causes the inference service to fail to start normally: Error: UNAVAILABLE: Internal: RuntimeError: (PreconditionNotMet) Tensor holds no memory. Call Tensor::mutable_data firstly When using a pdmodel format model weight, it is necessary to manually disable PIR optimization, and the inference service starts normally and the inference accuracy is generally satisfactory; As a hardware product under Baidu, when will it be improved and released to support PIR engine optimization?