Markus Hauru

Here's a snippet that shows the issue without using Turing.jl: ``` import Random import SequentialSamplingModels import Distributions import Optim import LinearAlgebra import StatsFuns import ForwardDiff Random.seed!(23) N = 100 true_params...

The above behavior seems to be a CPU specialty, on a TPU v2 donut it works fine. Should this perhaps be considered an XLA bug?

I guess the issue here is that `@docs f` is the natural syntax for both "all methods of `f`" and for "the generic, no-method `f`". I do find it non-optimal...

I certainly see the sense in which that is the most natural solution, but it is a heavily breaking one. How much that should weigh in the decision is not...

I would very much like to see this merged, `test_args` is a [standard part](https://pkgdocs.julialang.org/v1/api/#Pkg.test) of how `Pkg.test` operates, and would be great to have the Action support it. Could we...

>I am not sure how to specify step size for optimisation. @mhauru is that possible? Depends on the optimisation algorithm, but if the algorithm has a notion of a step...

@all-contributors please add @callummole for review

@all-contributors please add @radka-j for review

@all-contributors please add @AoifeHughes for ideas

@all-contributors please add @AoifeHughes for content