在rk3588上推理insightface模型只能使用cpu不能用npu加速

[Rsndmmm]

---

温馨提示：根据社区不完全统计，按照模板提问，可以加快回复和解决问题的速度

---

环境

• 【FastDeploy版本】：
• 【编译命令】rk3588按照https://github.com/PaddlePaddle/FastDeploy/blob/develop/docs/cn/build_and_install/rknpu2.md#%E7%BC%96%E8%AF%91fastdeploy-python-sdk编译
• 【系统平台】: Linux version 5.10.110 (root@jensen) (aarch64-linux-gcc (ctng-1.25.0-119g-FA) 11.3.0
• 【硬件】： friendly r6c
• 【编译语言】： Python 3.9

问题日志及出现问题的操作流程

• 附上详细的问题日志有助于快速定位分析
• 【模型跑不通】
  按照https://github.com/PaddlePaddle/FastDeploy/blob/develop/examples/vision/faceid/insightface/rknpu2/python/README_CN.md文档测试insightface，使用https://github.com/PaddlePaddle/FastDeploy/blob/develop/examples/vision/faceid/insightface/rknpu2/python/infer_arcface.py脚本，python infer_arcface.py --model ms1mv3_arcface_r100.onnx --face face_0.jpg --face_positive face_1.jpg --face_negative face_2.jpg --device npu 命令推理onnx模型时报错：
  [INFO] fastdeploy/runtime/runtime.cc(326)::CreateOrtBackend Runtime initialized with Backend::ORT in Device::CPU.
  [ERROR] fastdeploy/runtime/backends/ort/ort_backend.cc(401)::Infer Failed to Infer: Unexpected input data type. Actual: (tensor(uint8)) , expected: (tensor(float))
  [ERROR] fastdeploy/vision/faceid/contrib/insightface/base.cc(70)::BatchPredict Failed to inference by runtime.
  [ERROR] fastdeploy/runtime/backends/ort/ort_backend.cc(401)::Infer Failed to Infer: Unexpected input data type. Actual: (tensor(uint8)) , expected: (tensor(float))
  [ERROR] fastdeploy/vision/faceid/contrib/insightface/base.cc(70)::BatchPredict Failed to inference by runtime.
  [ERROR] fastdeploy/runtime/backends/ort/ort_backend.cc(401)::Infer Failed to Infer: Unexpected input data type. Actual: (tensor(uint8)) , expected: (tensor(float))
  [ERROR] fastdeploy/vision/faceid/contrib/insightface/base.cc(70)::BatchPredict Failed to inference by runtime.
  /home/pi/FastDeploy/examples/vision/faceid/insightface/python/infer_npu.py:12: RuntimeWarning: invalid value encountered in scalar divide
  return mul_ab / (mul_a * mul_b)
  FaceRecognitionResult: [Empty Result]FaceRecognitionResult: [Empty Result]FaceRecognitionResult: [Empty Result]Cosine 01: nan
  Cosine 02: nan
  RuntimeOption(
  backend : Backend.ORT
  cpu_thread_num : -1
  device : Device.CPU
  device_id : 0
  external_stream : None
  model_file : ms1mv3_arcface_r100.onnx
  model_format : ModelFormat.ONNX
  model_from_memory : False
  openvino_option : <fastdeploy.libs.fastdeploy_main.OpenVINOBackendOption object at 0x7f8c2e48f0>
  ort_option : <fastdeploy.libs.fastdeploy_main.OrtBackendOption object at 0x7f8c2e48f0>
  paddle_infer_option : <fastdeploy.libs.fastdeploy_main.PaddleBackendOption object at 0x7f8c2e48f0>
  paddle_lite_option : <fastdeploy.libs.fastdeploy_main.LiteBackendOption object at 0x7f8c2e48f0>
  params_file :
  poros_option : <fastdeploy.libs.fastdeploy_main.PorosBackendOption object at 0x7f8c2e48f0>
  trt_option : <fastdeploy.libs.fastdeploy_main.TrtBackendOption object at 0x7f8c2e48f0>
  )