TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Julia 1.6 Highlights

410 点作者 mbauman大约 4 年前

17 条评论

Buttons840大约 4 年前
I recently ported a reinforcement learning algorithm from PyTorch to Julia. I did my best to keep the implementations the same, with the same hyperparameters, network sizes, etc. I think I did a pretty good job because the performance was similar, solving the CartPole environment in the a similar number of steps, etc.<p>The Julia implementation ended up being about 2 to 3 times faster. I timed the core learning loops, the network evaluations and gradient calculations and applications, and PyTorch and Julia performed similar here. So it wasn&#x27;t that Julia was faster at learning. Instead it was all the in-between, all the &quot;book keeping&quot; in Python ended up being much faster in Julia, enough so that overall it was 2 to 3 times faster.<p>(I was training on a CPU though. Things may be different if you&#x27;re using a GPU, I don&#x27;t know.)
评论 #26655340 未加载
评论 #26590695 未加载
评论 #26589464 未加载
beeforpork大约 4 年前
Julia is such a wonderful language. There are many design decisions that I like, but most importantly to me, its ingenious idea of combining multiple dispatch with JIT compilation still leaves me in awe. It is such an elegant solution to achieving efficient multiple dispatch.<p>Thanks to everyone who is working on this language!
评论 #26590270 未加载
评论 #26584576 未加载
评论 #26584752 未加载
MisterBiggs大约 4 年前
I&#x27;ve been running the 1.6 release candidates, and the compilation speed improvements have been massive. There have been plenty of instances in the past where I&#x27;ve tried to &#x27;quickly&#x27; show off some Julia code, and I end up waiting ~45 seconds for a plot to show or a minute for a Pluto notebook to run, and that&#x27;s not to mention waiting for my imports to finish. It&#x27;s still slower than Matlab for the first run, but it&#x27;s at least in the same ballpark now.
评论 #26583780 未加载
评论 #26582860 未加载
snicker7大约 4 年前
On the package ecosystem side, 1.6 is required for JET.jl [0]. Despite being a dynamic language, the Julia compiler does a lot of static analysis (or &quot;abstract interpretation&quot; in Julia lingo). JET.jl exposes some of this to the user, opening a path for additional static analysis tools (or maybe even custom compilers).<p>[0]: <a href="https:&#x2F;&#x2F;github.com&#x2F;aviatesk&#x2F;JET.jl" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;aviatesk&#x2F;JET.jl</a>
评论 #26585726 未加载
评论 #26584848 未加载
cbkeller大约 4 年前
See also Lyndon’s blog post [1] about what all has changed since 1.0, for anyone who’s been away for a while.<p>[1] <a href="https:&#x2F;&#x2F;www.oxinabox.net&#x2F;2021&#x2F;02&#x2F;13&#x2F;Julia-1.6-what-has-changed-since-1.0.html" rel="nofollow">https:&#x2F;&#x2F;www.oxinabox.net&#x2F;2021&#x2F;02&#x2F;13&#x2F;Julia-1.6-what-has-chang...</a>
wiz21c大约 4 年前
Whatever improves loading times is more than welcome. It&#x27;s not really acceptable to wait because you import some libraries. In understand Julia makes lots of things under the hood and that there&#x27;s a price to pay for that but being a python user, it&#x27;s a bit inconvenient.<p>But I&#x27;ll sure give it a try because Julia hits a sweet spot between expressiveness and speed (at least for the kind of stuff I do : matrix, algorithms, graphs computations).
odipar大约 4 年前
I like Julia (mostly because of multiple dispatch). The only thing that&#x27;s lacking is an industry strength Garbage Collector, something that can be found in the JVM.<p>I know that you shouldn&#x27;t produce garbage, but I happen to like immutable data structures and those work better with optimised GCs.
评论 #26584206 未加载
评论 #26584354 未加载
评论 #26583787 未加载
noisy_boy大约 4 年前
How easy it is to produce a compiled executable in 1.6? I took a cursory look at the docs but couldn&#x27;t spot the steps for doing so.
评论 #26582256 未加载
评论 #26582478 未加载
评论 #26582601 未加载
systems大约 4 年前
I know its minor, but I still hope they will fix scoping<p>not that my suggestion is good, but what they have now is bad<p><a href="https:&#x2F;&#x2F;github.com&#x2F;JuliaLang&#x2F;julia&#x2F;issues&#x2F;37187" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;JuliaLang&#x2F;julia&#x2F;issues&#x2F;37187</a>
评论 #26583771 未加载
3JPLW大约 4 年前
The feature I&#x27;m most excited about is the parallel — and automatic — precompilation. Combined with the iterative latency improvements, Julia 1.6 has far fewer coffee breaks.
评论 #26581109 未加载
pjmlp大约 4 年前
Love the improvements, all those little details that improve the overall usability.
xiphias2大约 4 年前
Cool, I was thinking of downloading the RC, the demo was so impressive.<p>Will there be an M1 Mac version for 1.7?
评论 #26581412 未加载
fermienrico大约 4 年前
Are the performance claims of Julia greatly exaggerated?<p>Julia loses almost consistently to Go, Crystal, Nim, Rust, Kotlin, Python (PyPy, Numpy): <a href="https:&#x2F;&#x2F;github.com&#x2F;kostya&#x2F;benchmarks" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;kostya&#x2F;benchmarks</a><p>Is this because of bad typing or they didn&#x27;t use Julia properly in idiomatic manner?
评论 #26582695 未加载
评论 #26582450 未加载
评论 #26582508 未加载
评论 #26583928 未加载
评论 #26589628 未加载
评论 #26585042 未加载
评论 #26582386 未加载
评论 #26583969 未加载
评论 #26582587 未加载
JulianMorrison大约 4 年前
BTW, broken link on the documentation page, &quot;The documentation is also available in PDF format: julia-1.6.0.pdf.&quot; No it isn&#x27;t.
评论 #26585417 未加载
f6v大约 4 年前
Is there a per-project way to manage dependencies yet? I find global package installation to be the biggest weakness of all the R projects out there. Anaconda can help, but it’s not widely used for R projects. And Docker... well, don’t get me started.
评论 #26581553 未加载
评论 #26581558 未加载
评论 #26581544 未加载
评论 #26581581 未加载
siproprio大约 4 年前
On 1.6, I tried &quot;] add Plots&quot; and julia got stuck.
评论 #26585727 未加载
ng55QPSK大约 4 年前
maybe i misread this, but milestone &quot;1.6 blockers&quot; still has 3 open with &quot;1.6 now considered feature-complete. This milestone tracks release-blocking issues.&quot; - so how can 1.6 be ready?
评论 #26581867 未加载