[issue] Does anyone try TensorFlow Lite for Microcontrollers with OpenMV H7?

Special Interest Groups Micro
, ,
#1

Hi there,

I’m Leo, ML GDE from Mainland China. Months ago, I’ve tried tflite for microcontroller with OpenMV H7 but face some issues:

when converter.optimizations = [tf.lite.Optimize.DEFAULT] is set. it raises “hybrid model is not supposed” or converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8] and follow are set, it raises “Currently, only float32 input type is supported.”

more information can be found here

I’m wondering if anyone meets the same issue or can someone kindly sharing the possible solution?

thanks in advance :slight_smile:
Leo

#4

Hello, I am trying to run tensorflow model on OpenMV H7 Plus, I used tensorflow Object detection Api to train a model and it worked on my PC, I wonder how can I export and convert the model , thanks for your reply