I've been playing with #LMStudio for #localLLM with mediocre results. However #gemma4 really changed that. It's faster and is more capable then the other models I could try on my hardw…
@nelov@social.linux.pizza