![]() So on average we get for a treedepth >3 about of a ~35% speedup. The average speedup defined as parallel/serial is then by treedepth treedepth mean low_5% median high_95% The simulation code I wrote is attached if you want the details. In practice things are not that ideal such that, after discussing with we thought its good to first assess the merits of this approach using a simulation. This scheme works best if we have exactly alternating turns, of course. all checks still happen in the same order as before.then expand the fwd and back trajectories in full independence.pre-sample all the fwd and bck turns up to the maximal treedepth.I have been thinking about parallelizing our dynamic HMC sampler. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |