Latency Table when target hardware is not samsung note 10

[HayeonLee]

Hi, this work is awesome!

Latency table looks released only for Samsung note 10.

Could you release another latency tables especially for 1080ti?

Thanks in advnace.

Best, Hayeon Lee.

[Wheest]

I'm looking at this, and will probably roll my own at some point.

I guess I'm just curious if the authors answer what the "count" parameter in the LUT is, e.g. the provided yaml file here.

expanded_conv-input:7x7x192-output:7x7x320-expand:6-kernel:5-stride:1-idskip:0:
  count: 154
  mean: 5.264772727272727
  var: 0.19495998666150827

[tfwang08]

I'm looking at this, and will probably roll my own at some point.

I guess I'm just curious if the authors answer what the "count" parameter in the LUT is, e.g. the provided yaml file here.

expanded_conv-input:7x7x192-output:7x7x320-expand:6-kernel:5-stride:1-idskip:0:
  count: 154
  mean: 5.264772727272727
  var: 0.19495998666150827

I think "count" is number of measurements.Then calculate the mean and var.

[Jon-drugstore]

I'm looking at this, and will probably roll my own at some point. I guess I'm just curious if the authors answer what the "count" parameter in the LUT is, e.g. the provided yaml file here.

expanded_conv-input:7x7x192-output:7x7x320-expand:6-kernel:5-stride:1-idskip:0:
  count: 154
  mean: 5.264772727272727
  var: 0.19495998666150827

I think "count" is number of measurements.Then calculate the mean and var.

Do you have the details of latency model? how to build this latency table for own target hardware? Thanks for your reply!