How to achieve inference speed?

#5
by Light111 - opened

I am trying to do the inference for the same on a m1, it is taking about 15 seconds, on the nvdia t4 it is taking about 6-7 seconds, to achieve the mentioned (15 FPS), is there any need for some model optimisations like prunning or it will achieve the same without any optimisations?

Sign up or log in to comment