-
I am currently fitting a source to data, and while it worked fine yesterday, as of today I get an error when running the LogLikelihood.bestfit function. Below is the code I am running (more or less) a…
-
Trying out some autodifferentiation with Zygote I was surprised by code not being autodifferentiable because it made use of `@warn` macro.
Supporting try/catch in Zygote is[ an open issue](https://…
-
Hi! In the [website](https://www.kernel-operations.io/geomloss/api/pytorch-api.html) it is said: "full support of PyTorch’s autograd engine".
However, in the function [sinkhorn_loop](https://github…
-
On the documentation page [LV-Univ.md]( https://diffeqflux.sciml.ai/dev/examples/LV-Univ/), the introductory text states:
> Here's an example of doing this with both reverse-mode autodifferentiatio…
-
Hi, thanks for the awesome package! Unfortunately I'm having issues using it for my work. I basically want to integrate a set of equations of motions defined by a Hamiltonian `H(p, x)` that is too com…
-
I'd like `optimize` to work with dual numbers. This would be handy when optimizing a function that optimizes something.
Would it be difficult to write all the algorithms in a way that work with dual…
-
I am using JuliaApproximation to solve the Lane-Emden equation.
Following the example of Nonlinear Boundary Value problems for Solving a nonlinear boundary value problem satisfying the ODE with boun…
-
I work with larger array sizes, but have pre-existing python code to perform quick gradient calculations even with these arrays.
I was wondering if it was possible to "turn off" the auto-differentiat…
-
I'm considering inserting a weightless transform between CNN layers in a neural network. The transform is very complicated - details below. I'm aware that custom transformations are supported via e.g.…
-
https://github.com/IGITUGraz/eligibility_propagation/blob/7dd1558b5e2ff85b3b2b58e933cb2103a12f2a3b/Figure_4_and_5_ATARI/main.py#L235-L239
It seems the BPTT still be used here. How is e-prop used in…
ShuJz updated
4 years ago