privacy wins of local models: what are you seeing?

[vramstream]

I think the bottleneck is somewhere else. The reported numbers can look compelling, but in ordinary usage I didn't see the same benefit. Not saying yo

I hit almost the same issue last night.