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.

Sign up for GitHub

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

Jump to bottom

Add more engines supporting ChainRules to the docs #676

Merged
merged 3 commits into from
May 17, 2024
Merged

Add more engines supporting ChainRules to the docs #676

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
525723c
Add more engines supporting ChainRules to the docs
oxinabox May 17, 2024
e6c28cf
Add Tracker too
oxinabox May 17, 2024
6b390ff
Update docs/src/index.md
oxinabox May 17, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 5 additions & 1 deletion docs/src/index.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,10 +45,14 @@ It also `unthunk`s every tangent.

[ReversePropagation.jl](https://github.com/dpsanders/ReversePropagation.jl) is a reverse-mode AD that supports using `rrule`s for scalar functions, but not calling back into AD and opting out of rules.

[TaylorDiff.jl](https://github.com/JuliaDiff/TaylorDiff.jl) is a forward taylor mode AD.

### Packages supporting importing rules from ChainRules.
Several packages do not automatically load rules from ChainRules by default, but support importing rules that were defined using it, e.g. with a macro.

[ReverseDiff.jl](https://github.com/JuliaDiff/ReverseDiff.jl) is a reverse-mode AD that supports using `rrule`s, but not calling back into AD and opting out of rules. It requires opting in to each rule using a macro.
[ReverseDiff.jl](https://github.com/JuliaDiff/ReverseDiff.jl) is a reverse-mode AD that supports using `rrule`s, but not calling back into AD and opting out of rules.
[Tracker.jl](https://github.com/FluxML/Tracker.jl) is a reverse mode AD that supports importing `rrules`
[Enzyme.jl](https://github.com/EnzymeAD/Enzyme.jl) is a forward and reverse mode AD and supports import of `frule`s and `rrule`s.

[Tapir.jl](https://github.com/withbayes/Tapir.jl/) is a reverse-mode add that supports importing a restricted subset of rules defined using `rrule`. Specifically, rules for functions whose inputs have tangent type `Float64` or `NoTangent`.

Expand Down
Loading