-
I'm writing this issue to capture some of the ideas mentioned in a private discussion with @wsmoses about the future C++ interface. I'm hoping that this will increase visibility and allow other contri…
-
### Description
I'm working on a library for numerical quadrature in JAX, with derivatives defined via Leibniz rule. I've defined ``custom_jvp`` for my quadrature functions and it works fine in for…
-
Typst is a new markup-based typesetting system that is powerful and easy to learn.
* Project link: https://github.com/typst/typst
Typst doesn't have built-in support for diagrams yet. It would b…
-
Is there anything that I can do to help fix https://discourse.julialang.org/t/error-with-bunchkaufman-factorization-in-turing-jl/88174 ?
I don't have experience with the forward-diff functionality.…
-
For Task 1.2, I believe the intended code for `scalar.Scalar.__sub__`is `return Add.apply(self, -b)`, which passes the tests. Interestingly, if `return Add.apply(self, Neg.apply(b))` is used, the tes…
tnwei updated
2 years ago
-
For those not familiar with it, Kokkos (https://github.com/kokkos/kokkos) is a C++ library and programming model for portable, shared-memory parallelism. A program written once using Kokkos can be com…
-
## 🐛 Bug
In https://github.com/pytorch/audio/pull/1420, we are adding autograd check to `torchaudio.functional.phase_vocoder` and we noticed that it fails autograd test when `rate=0.7`, while it wo…
-
To compute gradients with Enzyme on a non-mutable array (like `SArray`), I need to annotate it as `Active`. Otherwise, if I do this
```julia
using StaticArrays, Enzyme
s = SVector(1.0, 2.0)
ds = zero(…
-
[Design doc](https://docs.google.com/document/d/1jbpGwAJLAfTAi6IZyPUSArgXt_Km49573UI7izsDvsE/edit#)
## Reading list
- https://popl21.sigplan.org/details/lafi-2021-papers/9/Decomposing-reverse-mo…
-
I just discovered the Diffrax package and it's great! However, I'm encountering an issue where the gradient evaluation is 40-80 times slower than the forward pass for my particular network (a potentia…