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
There are a few things missing from the := to reach peak utility:
Tracked variables are only included in calls to sample => not currently used in returned or predict, i.e. it's non-trivial to extract := values using a chain.
:= values are included in the chain returned from sample. However, it is treated as a standard parameter in the resulting chain! This means that it's not possible for downstream functionality, e.g. predict, to determine which parameters are actual parameters taht should go into, say, a VarInfo, and parameters that are from := expressions and thus should not go in, say, a VarInfo. Ref: Track := variables in resulting chains DynamicPPL.jl#693
Some things we'd like to track involve expensive computations that are unnecessary for inference. Hence we should ideally have a way to only activate this when we're not sampling. Ref: Convenience macros to use within @model DynamicPPL.jl#714
The text was updated successfully, but these errors were encountered:
There are a few things missing from the
:=
to reach peak utility:sample
=> not currently used inreturned
orpredict
, i.e. it's non-trivial to extract:=
values using achain
.:=
values are included in thechain
returned fromsample
. However, it is treated as a standard parameter in the resulting chain! This means that it's not possible for downstream functionality, e.g.predict
, to determine which parameters are actual parameters taht should go into, say, aVarInfo
, and parameters that are from:=
expressions and thus should not go in, say, aVarInfo
. Ref: Track:=
variables in resulting chains DynamicPPL.jl#693@model
DynamicPPL.jl#714The text was updated successfully, but these errors were encountered: