Stan on GPU: looking for model+dataset examples for empirical evaluation of speedups

General
8

I previously did this here. Though the dropbox file no longer exists I pasted the results into the conversation

EDIT: Also this was for the old implementation using ViennaCL, not our current implementation