Time inference is same after pruning model by Pytorch

[AouatifZ]

Hello

Please, I have a problem concerning the optimization of YOLOX model by Pytorch Pruning (Global & Local)
I got the same inference time after the optimization, I don't understand why ?

is it that the prune model by Pytoch does not remove the tensors which have equal to 0 after the equalization or there is another thing

any help

thanks in advance