Hi,
I've been using HiSim for LLM inference performance simulation, and I noticed that it currently seems to support only the H20 accelerator. Are there plans to add support for defining models or hardware configurations through a dictionary or configuration file?
regards.