I used to apply github code on my project. multiple model compression kernel is useful. but your benchmark tool (e.g. sparse kernel) is depennds on google benchmark library. it is cumbersome to use benchmark mulitple model. I hope you to provide end-to-end benchmark model like tensorflow benchmark model
I used to apply github code on my project. multiple model compression kernel is useful. but your benchmark tool (e.g. sparse kernel) is depennds on google benchmark library. it is cumbersome to use benchmark mulitple model. I hope you to provide end-to-end benchmark model like tensorflow benchmark model