Hi, I'm working on TinyML project for my work. for succesfully proceeding project, I apply your un-aligned sparse kernel on my project. but I have some trouble shooting to apply your sparse kernel. when I run on my sparse MobileNet v1 , which is applied by 4x1 simd aware pruning, there is no improvement in latency performance. How can I spect kernel used to my model. I use odroid xu4
Hi, I'm working on TinyML project for my work. for succesfully proceeding project, I apply your un-aligned sparse kernel on my project. but I have some trouble shooting to apply your sparse kernel. when I run on my sparse MobileNet v1 , which is applied by 4x1 simd aware pruning, there is no improvement in latency performance. How can I spect kernel used to my model. I use odroid xu4