Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

A bug with broadcasting, but whose? #10

Open
mcabbott opened this issue Mar 30, 2022 · 0 comments
Open

A bug with broadcasting, but whose? #10

mcabbott opened this issue Mar 30, 2022 · 0 comments

Comments

@mcabbott
Copy link

While replying to this https://discourse.julialang.org/t/ann-forwarddiffpullbacks-jl-forwarddiff-based-chainrulescore-pullbacks/78737/10 I got an error in the following example:

julia> using ForwardDiffPullbacks, Zygote

julia> Zygote.gradient([1,2,3], 4) do xs, y
         f3 = x -> abs2(@show(x)/y)
         sum(fwddiff(f3).(xs))  # this cannot track gradient w.r.t. y
       end
x = 1
x = 2
x = 3
x = Dual{ForwardDiff.Tag{Tuple{var"#84#86"{Int64}, Val{1}}, Int64}}(1,1)
x = Dual{ForwardDiff.Tag{Tuple{var"#84#86"{Int64}, Val{1}}, Int64}}(2,1)
x = Dual{ForwardDiff.Tag{Tuple{var"#84#86"{Int64}, Val{1}}, Int64}}(3,1)
([0.125, 0.25, 0.375], nothing)

julia> Zygote.gradient([1,2,3], 4) do xs, y
         f3 = x -> abs2(@show(x)/y)
         sum(f3.(xs))  # reverts to slower generic broadcast, no Dual
       end
x = 1
x = 2
x = 3
([0.125, 0.25, 0.375], -0.4375)

julia> f4(x,y) = abs2(@show(x)/y);

julia> Zygote.gradient((xs,y) -> sum(f4.(xs, y)), [1,2,3], 4)
x = Dual{Nothing}(1,1,0)
x = Dual{Nothing}(2,1,0)
x = Dual{Nothing}(3,1,0)
([0.125, 0.25, 0.375], -0.4375)

julia> Zygote.gradient((xs,y) -> sum(fwddiff(f4).(xs, y)), [1,2,3], 4)
x = 1
x = 2
x = 3
x = Dual{ForwardDiff.Tag{Tuple{typeof(f4), Val{1}}, Int64}}(1,1)
x = Dual{ForwardDiff.Tag{Tuple{typeof(f4), Val{1}}, Int64}}(2,1)
x = Dual{ForwardDiff.Tag{Tuple{typeof(f4), Val{1}}, Int64}}(3,1)
x = 1
x = 2
x = 3
ERROR: MethodError: no method matching (::ChainRulesCore.ProjectTo{Float64, NamedTuple{(), Tuple{}}})(::Vector{Float64})
Closest candidates are:
  (::ChainRulesCore.ProjectTo{T})(::AbstractFloat) where T<:AbstractFloat at ~/.julia/packages/ChainRulesCore/RbX5a/src/projection.jl:171
  (::ChainRulesCore.ProjectTo{<:Number})(::ChainRulesCore.Tangent{<:Complex}) at ~/.julia/packages/ChainRulesCore/RbX5a/src/projection.jl:192
  (::ChainRulesCore.ProjectTo{<:Number})(::ChainRulesCore.Tangent{<:Number}) at ~/.julia/packages/ChainRulesCore/RbX5a/src/projection.jl:193
  ...
Stacktrace:
 [1] _project
   @ ~/.julia/packages/Zygote/H6vD3/src/compiler/chainrules.jl:182 [inlined]
 [2] map(f::typeof(Zygote._project), t::Tuple{Vector{Int64}, Int64}, s::Tuple{Vector{Float64}, Vector{Float64}})
   @ Base ./tuple.jl:247
 [3] gradient(::Function, ::Vector{Int64}, ::Vararg{Any})
   @ Zygote ~/.julia/packages/Zygote/H6vD3/src/compiler/interface.jl:77

The bug may well be in Zygote, or ChainRulesCore, and thus my fault... but I open this somewhere so as not to forget.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant