💎 @googlegemma

@xeophon @lmstudio @GoogleDeepMind Why?

@xeophon @lmstudio @GoogleDeepMind what are ur usecases? "rewrite", "summarize", "translate," or something bigger in scope and harder by nature?

@xeophon @lmstudio @GoogleDeepMind yes, even i have one model always loaded on my system for assistance while building stuff or solving any problems.

i think people who can use small 4-9B models to build stuff can actually be called coders.

@xeophon @lmstudio @GoogleDeepMind mac specs?

@xeophon @lmstudio @GoogleDeepMind Are you using it for the privacy considerations, Xeo?

@xeophon @lmstudio @GoogleDeepMind What context window are you using?

@wambosec @lmstudio @GoogleDeepMind M4 Max + 64 GB RAM

@JeremyNguyenPhD @lmstudio @GoogleDeepMind Latency

@dgreller @lmstudio @GoogleDeepMind 4K, but for my use cases I can prob go as low as 1K. I got a good Mac, though.

@xeophon @lmstudio @GoogleDeepMind E2B is also great model for RL experiments

@Laz4rz @lmstudio @GoogleDeepMind Cause it’s good

@stalkermustang @lmstudio @GoogleDeepMind Basically this, yeah. That’s where local models are useful and win in latency

@xeophon @lmstudio @GoogleDeepMind Woaw

@xeophon @lmstudio @GoogleDeepMind even 12b model is worth trying and it literally gives gpt4o types vibe

@xeophon @lmstudio @GoogleDeepMind people who can use small 4-9B models to get assistance for building stuff*