Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
byt143
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
31.
▲
by
byt143
7y ago
To which version of Julia are you referring? All of those things are much improved now.
32.
▲
by
byt143
7y ago
Even in swift it isn't "turtles all the way down". The AD stuff is hardcored into the C++ guts of the compiler, whereas Julia's source to source autodiff accesses a compiler pass from a fully Julia user package. Aside fr
33.
▲
by
byt143
7y ago
Would you be open to doing a Fast.ai initiative in Julia? If so, what would that take?
34.
▲
On the performance and design of BioSequences compared to the Seq language
(biojulia.net)
3 points
by
byt143
7y ago
|
0 comments
35.
▲
by
byt143
7y ago
Not sure the data supports that ,: https://juliacomputing.com/blog/2020/01/06/january-newslette...
36.
▲
by
byt143
7y ago
Better static compilation and compile times which will reduce the time to first plot.
37.
▲
by
byt143
7y ago
Specifically what?
38.
▲
by
byt143
7y ago
This sounds amazing and exactly what people have been clamoring for. Extremely excited. I think it will also rocket up Julia's popularity and usage not just for ML but as a general purpose language
39.
▲
by
byt143
7y ago
Julia,'s source to source autodiff is implemented as a custom compiler pass in a third party package. Similar approached are bring taken by prob programming packages.
40.
▲
by
byt143
7y ago
Yes it does. See genie.jl for a full mvc framework. Also there's mux.jl and http.jl.
41.
▲
by
byt143
7y ago
It's already bring done in Julia. See zygote.jl. Works with typed IR equivalent to a core compiler pass but from a third party package with regular Julia staged programming.
42.
▲
by
byt143
7y ago
It's on Jeff's radar: https://www.youtube.com/watch?v=TPuJsgyu87U
43.
▲
by
byt143
7y ago
http://genieframework.com/
44.
▲
by
byt143
7y ago
This really showcases Julia's power. This is a DSL built on top of Julia, written in pure Julia. So, a Julia package ;) https://github.com/probcomp/Gen
45.
▲
by
byt143
7y ago
What do you see as Julia's most egregious problems, internal or otherwise? and internal in particular?
46.
▲
by
byt143
7y ago
This goes way way beyond graphical models.
47.
▲
by
byt143
7y ago
>The syntax doesn't seem to be changed compared to Julia (so is it actually a new language or just a library? And the semantics are the same with some additions. This is a good thing as it allow interop with Julia code, low mental o
48.
▲
by
byt143
7y ago
I had the same thoughts upon seeing the list. Luckily package compilation seems is on the compiler dev roadmap. Traits are more of a 2.0 thing I think.
49.
▲
by
byt143
7y ago
So many patently false claims here. First, julia has it's own IR (comparable to swift IR as mentioned by Chris), which is used to great effect in the autodiff packages to allow for compile time zero overhead gradients. The flexibility
50.
▲
Beyond Machine Learning Pipelines
(julialang.org)
2 points
by
byt143
7y ago
|
0 comments
51.
▲
by
byt143
8y ago
Worse than python?
52.
▲
by
byt143
8y ago
Yes. see this as an example: http://genieframework.com/
53.
▲
by
byt143
8y ago
"these third party libraries are using, like reliance on iterator protocol, descriptor protocol, metaclasses, context managers, Python data model special methods, dynamic attribute lookups, etc.," Julia can Jit code that relies on
54.
▲
by
byt143
8y ago
"Here is an easy way to prove me wrong: Find the minimum of a pure python function quickly, with whatever python tool you want, without having to deal with the enormous python function call overhead, without rewriting the function. Ima
55.
▲
Rock–paper–scissors game in less than 10 lines of code
(giordano.github.io)
3 points
by
byt143
8y ago
|
0 comments
56.
▲
by
byt143
8y ago
Both cassette based performance linting and static type checking will solve this problem. There are already the beginnings of the tooling being formed.
57.
▲
Queryverse: Julia Packages for Data Science
(queryverse.org)
3 points
by
byt143
8y ago
|
0 comments
58.
▲
by
byt143
8y ago
It's possible to have both the dynamism and static type checking in Julia. Check out this issue: https://github.com/FluxML/Flux.jl/issues/614
59.
▲
by
byt143
8y ago
Why is it a silly insistence? What about it is not fit for general programming? Please be specific.
60.
▲
by
byt143
8y ago
Re: Debugger, fair enough but it's in the works Re: Scoping rules, these are being evaled. Re; startup time, already better in 1.1, and will soon be marginalized from two ends: Better static compilation and better tiered compilation.
More ›