You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A good point raised about helping the user with the choice of the AD backend is that a lot of users don't really want to know anything about autodiff, etc. but just want to get their inference done (a la Stan).
One suggestion brought up by @SamuelBrand1 is that it would be helpful if we just had a warning for the user in cases such as if dimension is >15.
An alterantive might be to introduce a automatic_ad_choice(model) which just benchmarks it for different AD backends (and checks correctness), and then we just tell the user to call this for perf.
Both of these requires some thought as to exactly how to implement this though 😬
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
A good point raised about helping the user with the choice of the AD backend is that a lot of users don't really want to know anything about autodiff, etc. but just want to get their inference done (a la Stan).
One suggestion brought up by @SamuelBrand1 is that it would be helpful if we just had a warning for the user in cases such as if dimension is >15.
An alterantive might be to introduce a
automatic_ad_choice(model)
which just benchmarks it for different AD backends (and checks correctness), and then we just tell the user to call this for perf.Both of these requires some thought as to exactly how to implement this though 😬
@yebai @sunxd3 @mhauru @penelopeysm
Beta Was this translation helpful? Give feedback.
All reactions