-
@willtebbutt @oxinabox
I'd like suggestions/advice on how we can get autodiff support for `LoopVectorization.@avx`.
Where / how can we integrate things to get things to work with the various AD to…
-
@dpsanders has just released [ReversePropagation.jl](https://github.com/dpsanders/ReversePropagation.jl) which is a tape-to-source scalar reverse mode AD using ChainRules.
We have reverse depenency …
-
We need integration testing github workflows like the ones in ChainRules and DynamicPPL to avoid breakages in TuringLang and SciML at least.
-
```julia
julia> gradient(x -> evalpoly(x, (1,2,3.5)), 2.0)
(16.0,)
julia> gradient(x -> evalpoly(x, [1,2,3.5]), 2.0)
(16.0,)
julia> gradient(x -> sum(evalpoly.(x, Ref((1,2,3.5)))), [2.0])
([…
-
Currently, implementations don't support automatic differentiation with packages such as `ForwardDiff.jl`.
Probably best to give ChainRules such as in https://github.com/JuliaMath/SpecialFunctions…
-
v1 is the standard basis vector
```
julia> Zygote.gradient(v->0v, v1)
ERROR: cannot convert from ⟨×××⟩ to ⟨+++---⟩*
Stacktrace:
[1] error(s::String)
@ Base ./error.jl:33
[2] (::SubMan…
-
## WIP PRs
- [ ] ~#665 -- Attempts to compile Lux models to XLA. While more generally scoped to allow models that cannot be compiled via Reactant, this is particularly hard because we need to give ow…
-
The most recent iteration of this discussion was https://github.com/FluxML/NNlib.jl/pull/434, but it goes back to https://github.com/JuliaDiff/ChainRulesCore.jl/issues/547 and much further. Given that…
-
OS: WSL2 Ubuntu 20.04 - on Windows 10 Pro
`julia 1.7.0-rc1` build from github repo on Ubuntu 20.04
```
julia > versioninfo()
Julia Version 1.7.0-rc1
Commit 9eade6195e (2021-09-12 06:45 UTC)
…
-
This might not be a bug, but at the least it seems to be a mysterious error.
I'm using Flux and in the loss function (which is processed by Zygote and so uses ChainRules.jl), I'm using `prod`. The …