You just sample from a grammar and you automatically get 100%; who knows but it seems the most likely thing they are doing. llama.cpp has supported this for a while ( using a BNF-style grammar -- https://github.com/ggerganov/llama.cpp/blob/master/grammars/... )
edit: oh actually, we do sort of know -- they call out jsonformer as an inspiration in the acknowledgements
edit: oh actually, we do sort of know -- they call out jsonformer as an inspiration in the acknowledgements
https://github.com/1rgs/jsonformer