-
There are a bunch of autodiff functors that are implemented in Stan but not exposed yet in BridgeStan. The two most basic are already done. Most of them other than directional derivatives require f…
-
Hi,
I tried to use AutoEnzyme as an optimizer to build a neural network to predict parameters for ODEs following the example of [DiffEqFlux](https://docs.sciml.ai/DiffEqFlux/stable/examples/neural_od…
-
Thank you for a great package!
I would like to use Zygote on `quadgk` but below is a self-contained example, where it does not work because `quadgk` calls `setindex!`. Can I do anything to work arou…
-
avhz updated
4 months ago
-
* Implement forward-mode AD for temporal connections; reverse-mode AD for layer connections
* Update at each time step
* Approximate BPTT without having to store any historical gradients
-
I encountered unexpected behavior with a function and I am unsure what the issue is (compiled via `cargo +enzyme run --release`)
```rust
// this crashes
#[autodiff(df2f, Forward, Const, Dual, D…
-
@benjaminhuth made this whole thing work with autodiff. It would be cool if we added some CI tests that built against it. It shouldn't be a strict dependency though, so we'd have to have a flag to avo…
-
Hi!
As you know, @ExpandingMan and I are looking to optimize performance for StaticArrays. Forward mode works splendidly, but reverse mode still makes one allocation during the `gradient` call:
```jul…
-
There are problems with the Graphviz plotting of the computation graphs in `autodiff`:
- [ ] The nodes don't indicate the operation that took place (e.g. `+`, `sin`).
- [ ] There are double links …
avhz updated
11 months ago
-
I tried running the last example in the README with an HMC sampler substituted for the MALA one:
```
using Klara
plogtarget(z) = -dot(z, z)
p = BasicContMuvParameter(:p, logtarget=plogtarget…