Run customized arm-nn model

[mahesh93]

Hi,

  I have one basic question related to enabling eIQ on imx8xqxp b0 board,

  I could run demo example but getting weird numbers on GPU, 

  Ideally with GPU execution time should be less but I am getting more here.

------------------------------------------------------------------------------

So going ahead i have some questions:

1. how to run customized arm-nn models on this board ?

   I have caffemodel, which armnn_binary binary to use ? where i will get that ?

2. So to see the GPU is used ? is there any way we can check computation on GPU ?

Thanks