-
LabelledArrays looks very promising to simplify model definitions. Unfortunately it seem largely incompatible with AD. Is there a good workaround?
```Julia
using LabelledArrays
import ForwardDiff
…
-
#### Summary:
I have a small example program that dies with "Rejecting initial value: Gradient evaluated at the initial value is not finite" even though the log probability has a finite gradient ever…
-
| | |
|------------------|-----------------|
|Previous ID | SR-14115 |
|Radar | None |
|Original Reporter | @dan-zheng …
-
e.g.
```julia
julia> using TropicalNumbers, Octavian, TropicalGEMM, BenchmarkTools, Tullio
julia> v2!(r, d) = @tullio (max) r[j, i] = d[k, j] + d[i, k]
v2! (generic function with 1 method)
ju…
-
Currently, the DiscreteSpace is defined as `{1, ..., n}` (as it should be), but the lines in `CartPole.jl` that map `{1, 2} --> {-1, 1}` are commented out. Additionally, the assertion is commented out…
-
Matrix3x3 is faster because it will have no heap allocation and no TensorFlow overhead.
As mentioned in https://github.com/borglab/SwiftFusion/issues/42#issuecomment-627413240, this is currently th…
-
XAD should support [Eigen](https://eigen.tuxfamily.org/) data types and related linear algebra operations, calculating values and derivatives efficiently.
Ideally, simply using an XAD type within E…
-
Looks like tangent only supports autograd on operations registered at `adjoint`. Is it a good idea to simply ignore those non-registered operations when doing autograd?
```python
def f(x):
a = …
-
Currently tensors are immutable, though may share the same underlying data.
We do have some *localized* mutation through `.reverseDiff()` that creates a *new* tensor with a mutable register cha…
dsyme updated
4 years ago
-
### Background
[This issue](https://github.com/FluxML/Zygote.jl/pull/445) does stuff with `FillArrays`, and got be thinking about a problem we've known about for a while but not successfully resolv…