Deep learning model simulation

[bdipta01]

I wrote a deep learning model using C language. When I tried to simulate using verilator, it stuck for 10 hours. Then I cancelled the simulation. I used Gemmini config. How can I make it faster?
But when I simulated using qemu and spike, it took 1 min and 13 mins respectively.
I need to check the deep learning model performance by changing the core number, cache number, and other configuration. How can I do that?