Deployment File on raspberry Pi4

#1

Good morning,

I hope this email finds you in good health.

I am trying to deploy a model on a Raspberry Pi 4. However, when I specify the target and compile the program, I encounter this result. Have you, or someone you know, encountered this problem before?

Thank you
problem :
Target results for float32:

{
“device”: “raspberry-pi-4”,
“tfliteFileSizeBytes”: 280312,
“isSupportedOnMcu”: false,
“timePerInferenceMs”: 2,
“mcuSupportError”: “Unsupported ops: FlexTensorListReserve, FlexTensorListStack, FlexTensorListSetItem.”
}

#2

Hi @Abdeljelil

Can you share the project ID??

Best

Eoin

We may be able to add the unsupported operators but I’m not sure how you got to this error. Please share all of the detail you can on your pi4 setup, did you follow the dependency step instructions?

Unsupported operators on linux based platforms:

If your model contains unsupported operations such as FlexTensorListReserve and FlexTensorListStack, it means these operations require flex delegates to run. To resolve this, you will need to install the flex delegates shared library on your Linux system. Follow these steps to install the flex delegate library:

  1. Download the shared library for your target architecture and operating system. The options include:

  • macOS, x86 (also runs on M1/M2 using Rosetta)

  • Linux, armv7 (most 32-bits Arm-based Linux systems, e.g., Raspberry Pi 4 running 32-bits Raspbian)

  • Linux, aarch64 (most 64-bits Arm-based Linux systems, e.g., Jetson Nano)

  • Linux, x86_64 (Intel/AMD based Linux systems)

  1. Place the libtensorflowlite_flex_2.6.5.so (or .dylib on macOS) file in /usr/lib or /usr/local/lib.