IMX 8M Plus, Runtime Error on making tflite-runtime interpreter with delegate (libvx_delegate.so)

取消
显示结果 
显示  仅  | 搜索替代 
您的意思是: 

IMX 8M Plus, Runtime Error on making tflite-runtime interpreter with delegate (libvx_delegate.so)

993 次查看
kandarp_rastey
Contributor II

IMX 8M Plus, Runtime Error on making tflite-runtime interpreter with delegate (libvx_delegate.so)

Platform = NXP-imx8mp-lpddr4-evk

Linux Kernel Release = 5.15.32-lts-next+gfa6c3168595c

 

made an efficientnet model using:

 

from tensorflow.keras.applications import EfficientNetB0
model = EfficientNetB0(weights='imagenet')

 

- Then converted the same to int8-tflite format to be able to run it on the NPU, while trying to run the same model on npu I got this strange runtime error with no details :

 

Traceback (most recent call last):
  File "/home/root/xyz/tflite_detection_live.py", line 143, in <module>
    main()
  File "/home/root/xyz/tflite_detection_live.py", line 78, in main
    c_interpreter = make_interpreter(args.classifier_model, args.use_npu)
  File "/home/root/xyz/tflite_detection_live.py", line 22, in make_interpreter
    return Interpreter(
  File "/usr/lib/python3.10/site-packages/tflite_runtime/interpreter.py", line 496, in __init__
    self._interpreter.ModifyGraphWithDelegate(
RuntimeError

 

 Note : This same model runs fine on the CPU.

- To check if there was something wrong in the conversion process, I tried to do the same with keras.applications.mobilenet_v2 and that model ran fine on NPU as well as the CPU.

 

- Looks like there can be some issues with the "libvx_delegate.so" being used.

- Any pointers as to how to run this efficientnet model on the NPU?

0 项奖励
回复
2 回复数

953 次查看
kandarp_rastey
Contributor II

- yes, tried to run the same on a variscite imx8mplus evk with kernel =  5.15.60-imx8mp+g15e390fe9060.

- But, got the same runtime error.

0 项奖励
回复

970 次查看
Bio_TICFSL
NXP TechSupport
NXP TechSupport

Hello,

 

Please try the latest bsp 5.15.71, since your kernel version is not supported, you may try the latest one.

Regards

0 项奖励
回复