Naveen-Dodda

@jk78346 Have you tried optimize the model with edgetpu_compiler 15.0. let us know if the path is still being explored.

@wwooster I have few questions to understand your end goal here. so you want to perform matrix multiplication on 3000 matrices each 500x500 size ? your input =[1, 3000,500,500] output...

@wwooster If it is a Deep learning model all the weights are fused and model would technically have fixed set of matrices on with which input is modified. But in...

@Spanni26 Not exactly sure how to reproduce this issue on different x86_64 platform. But these are docker builds things should fairly straight forward if you use docker. Can you follow...