-
I am having trouble using DynamicHMC + Zygote on a simple example using a constrained variable. This will work with ForwardDiff, but I am interested in Zygote because of the reverse mode autodiff and …
-
For fitting parameter values for ODEs a la the adjoint sensitivity method, we might want to override the gradient computation for the forward ODE solve. More concretely, we might have an integrator fu…
-
This was opened as a response to the discovery in [this](https://github.com/FluxML/Zygote.jl/pull/366/files#diff-fbbb811698e867bdef17d2008dc4a8acR26-R27) Zygote PR that our handling of complex functio…
-
JAX supports custom primitives and vjps, just like Autograd did. Improvements:
1) add this to documentation
2) add a minimal example of this in the examples section
3) add a wrapper function if app…
-
JAX on master (1bb9aaa88c09a90a115c0304c05b0fc25932523b) is perfectly content letting one write
```python
vmap(lambda y0: odeint(f, y0, timesteps))(y0_vec)
```
But it's not immediately clear to me…
-
I've gotten stuck at a similar junction a few times. It's where I have a loss function that I'm feeding some stateful data structure, and I want the state of the data as well as the cost at the end. H…
-
## Summary:
Flatten the directory structure by getting rid of the `scal`, `arr`, `mat` folders. The goal is to make the Math library easier to navigate for developers and users.
## Description:
…
-
I am doing an ode computation that I want to vectorize using vmap. I need gradients. This all works fine until I add vmap --- then I get an AssertionError.
Here is a stripped-down script to reprodu…
-
#### Summary:
Support solving a system of non-linear equations using one of the better supported unsupported modules in Eigen. An alternative would be to use KINSOL in Sundials, but I think Eigen wou…
-
Can you please inform whether there are any plans to develop Calculus Ipython Notebook?