Process for building custom a model, training it, and deploying it to the OpenMV Cam H7

OpenMV Products
21

Might want to double check the model is not corrupted, if you reset the camera after copying a file without safe removing the disk (eject/unmount) files get corrupted.

22

Hi,
The gray-line error says: Input model data type should be 8-bit quantized!
However I have performed the quantization by following Quantização pós-treinamento  |  TensorFlow Lite It quantizes the model to 8 bits.
Thanks.

23

Can you verify that your input is an 8 bit unit or int and output is an 8 bit unit or int? This is probably not the case if you are getting that error.