Quantitative models are slower than the original models

[leizhu1989]

hello, when I try in c++ project to infer the Quantized models，I find it is slower than original float32 models. why is it?

[neuxys]

c++ project ? Do you rewrite the function of "def callback(indata, outdata, frames, buftime, status)" using c++ ?

[StuartIanNaylor]

Running the tflite runtime the quantised models are much faster on the intended Pi hardware.