If building ML kit with ARM Compiler 6.22 (6.21 works ok, probably is not just that strict as 6.22), build fails with a following error (there are couple of more same errors, so listing just one in here):
tensorflow/lite/kernels/internal/quantization_util.cc:202:14: error: use of infinity is undefined behavior due to the currently enabled floating-point options [-Werror,-Wnan-infinity-disabled]
202 | return std::numeric_limits::infinity();
I can quickly workaround this by disabling -Werror from tensorflow Makefile.
Br,
Kimmo
