Can you explain the concept behind each of the main size ranges of LLM models, as in, what hardware setups the different size niches are meant to fit into (~30b, ~70b, ~120b, ~230b, etc). Like is it mainly based on pro hardware sizing for 8-bit, or consumer GPU vram for ~Q4, or some mixture?

I am curious about intended sizings of the main size niches of the popular local LLM models.

As in, we can see there is a major niche at 26b-35b, then hardly anything from 36 through 69b, then (formerly) another major niche at ~70b-72b, then another niche at ~120b-123b, then another big gap till ~230b-235b, and then it gets a bit more mixed all over the place after that with 300b-750b being scattered more randomly probably based more on just whatever the best strength per size they could get when training the model of whatever it worked out to, rather than trying to force it into a specific size-niche of some sort, although maybe still a little bit of size nudging to get under some key size cutoffs of various sorts to do with server level hardware.

Anyway, for the noobs, can you explain the concept behind the different size ranges, for the more blatant ones around ~30b, ~70b, ~120b, and ~230b of what they are basing it on, like if it is to do with certain server hardware memory sizes, or prosumer/consumer hardware sizes, and at what quantization/bit levels.

I want to get a better sense for how these things are sized