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
If TaylorDiff fixed its stuff and stops yieldings NaNs on Infs, we might be able to switch to it.
The produced code would clearly be much faster, with the downside of a bit longer loading/precompilation time.
Dont forget to lift up the d constants in the type system so that it is type stable when we do that.
If TaylorDiff fixed its stuff and stops yieldings NaNs on Infs, we might be able to switch to it.
The produced code would clearly be much faster, with the downside of a bit longer loading/precompilation time.
Dont forget to lift up the
d
constants in the type system so that it is type stable when we do that.Still a few bugs however: See JuliaDiff/TaylorDiff.jl#61 for upstream report of the bugs.
The text was updated successfully, but these errors were encountered: