-
Larger PDE solves may benefit from a reverse mode AD and also helps with adding support for neural networks as surrogates or part of training the PDEs themselves, therefore I was wondering what it wou…
-
Currently, histograms (and some other types of plots?) can be used as adjoints in layouts, as in the home page example:
![holohome](https://cloud.githubusercontent.com/assets/1695496/11792952/88bcb54…
-
@ChrisRackauckas have you seen this work by Ricky already?
https://arxiv.org/abs/2009.09457
**"Hey, that's not an ODE": Faster ODE Adjoints with 12 Lines of Code**
Patrick Kidger, Ricky T. Q. C…
-
Currently we wrap NaiveBayes.jl, but only allow tabular input (internally converted to matrix) which limits application to NLP and elsewhere. However NaiveBayes.jl itself supports dictionary input.
…
-
For discretization of non-Hermitian operators in non-uniform finite-differences, it can be advantageous to work with separate left- and right-vectors, where the inner product is computed as `transpose…
jagot updated
3 years ago
-
The following bug turned up in #40504:
```julia
julia> using LinearAlgebra
julia> U = UpperTriangular(reshape(collect(3:27.0),5,5));
julia> B = reshape(collect(1:14.0),2,7); B2 = copy(B); B3 = cop…
-
It'd be great to have a neural machine translation model incorporated into our models repository.
saeta updated
4 years ago
-
Is there any reason we implement pretty much the whole functionality for typed comprehensions in lowering? Compare:
```julia
julia> Meta.@lower Int[i for i in 1:3]
:($(Expr(:thunk, CodeInfo(
…
-
Hi,
Julia's base `round` function takes a number and outputs the nearest integer. I want to customly define its gradient as the identity. I used Zygote's `@adjoint` macro very simply,
```julia
@a…
-
Discussion about new repo that automatically creates theano ops useable in pymc3 from python fenics functions [here](https://discourse.pymc.io/t/using-fenics-pde-solver-together-with-pymc3/5620/2). Li…