Hi Stefan,
I'm really pleased with the model, it works great on Ubuntu 20.04.
Now I am trying to install BirdNET-Lite on a RPi. This is using this part of the import "try" block:
import tflite_runtime.interpreter as tflite
Since a Tensorflow package isn't easily available for the platform, I have python3-tflite-runtime installed (2.5.0). Python 3.7 also.
When I execute the script, it fails with the following error:
LOADING TF LITE MODEL... Traceback (most recent call last):
File "analyze.py", line 238, in
main()
File "analyze.py", line 215, in main
interpreter = loadModel()
File "analyze.py", line 46, in loadModel
interpreter.invoke()
File "/usr/lib/python3/dist-packages/tflite_runtime/interpreter.py", line 540, in invoke
self._interpreter.Invoke()
RuntimeError: Regular TensorFlow ops are not supported by this interpreter. Make sure you apply/link the Flex delegate before inference.Node number 29 (FlexRFFT) failed to prepare.
For some reason it doesn't apply/link the Flex delegate as it does when using:
from tensorflow import lite as tflite
Hi Stefan, I'm really pleased with the model, it works great on Ubuntu 20.04.
Now I am trying to install BirdNET-Lite on a RPi. This is using this part of the import "try" block:
import tflite_runtime.interpreter as tflite
Since a Tensorflow package isn't easily available for the platform, I have python3-tflite-runtime installed (2.5.0). Python 3.7 also.
When I execute the script, it fails with the following error:
LOADING TF LITE MODEL... Traceback (most recent call last): File "analyze.py", line 238, in
main()
File "analyze.py", line 215, in main
interpreter = loadModel()
File "analyze.py", line 46, in loadModel
interpreter.invoke()
File "/usr/lib/python3/dist-packages/tflite_runtime/interpreter.py", line 540, in invoke
self._interpreter.Invoke()
RuntimeError: Regular TensorFlow ops are not supported by this interpreter. Make sure you apply/link the Flex delegate before inference.Node number 29 (FlexRFFT) failed to prepare.
For some reason it doesn't apply/link the Flex delegate as it does when using:
from tensorflow import lite as tflite
Any ideas why?
Thanks!