Zenna Tavares

Precompilation fails on `@agent`

3

**Describe the bug** Precompilation fails on simple examples which uses `@agent Zombie OSMAgent` **Minimal Working Example** 1. Generate a new package `TestAgents` 2. Replace TestAgents.jl with: ```julia module TestAgents using...

I understand the solver is random. But is there a way to fix the random seed?

Fails on shufflenet

1

ONNX's fails on Shufflenet, a network in the ONNX ModelZoo: https://github.com/onnx/models/tree/master/vision/classification/shufflenet ONNX.jl converts it to a Flux model fine. However, It fails (on an input of size (224, 224, 3,...

ONNX.jl fails to convert squeezenet (on input of size (244, 244, 3, 1)) https://github.com/onnx/models/tree/master/vision/classification/squeezenet Everything is fine until the reshape of the last layer. For some reason the reshape size...

Confusion with using optim with autograd

5

Thanks for the hard work on this, cool stuff. I'm trying to use optim with autograd, but from looking at the examples, and previous issues I'm left a little confused....

Lint doesn't support declarations on function return types

1

```julia julia> lintstr("f(x::Bool)::Bool=!x") 2-element Array{Lint.LintMessage,1}: none:1 E321 x: use of undeclared symbol none:1 I171 f(x::Bool): LHS in assignment not understood by Lint ```

Enable graphs with different Vertex and Edge types

4

In working on a constraint solving package, I wanted to use Graphs.jl to create graphs with different types of Vertices. I attempted to do this using an abstract Vertex type,...

I am not sure who if any one is at fault here, but when I use FastAnonymous with Distributions.jl (which uses DataArray) I get a slew of ambiguity errors ```...

``` julia> @Solve(x == 4, x) Symbol("$Aborted") ```

``` julia> function f(x) if x > 0 k = 0 while k < 10 k += 1 x = x * 4 end else j = 0 while j...