All ML really happens in C++. The only advantage that python has is that people use it. P6 is flexible enough that I'm sure you can port TensorFlow using nativeCall within a month. Perl as a whole has a terrible image problem. People think it's just awk with more punctuation variables. Literally, to most people Perl is regexes. In my opinion, what we need is a tutorial website that ranks high on search engines for searches not about Perl, so that people will realize that Perl is very readable and very useful. But I digress.
On Tue, Aug 13, 2019, 05:18 Francis Grizzly Smit <griz...@smit.id.au> wrote: > > On 13/08/2019 12:03, Eliza wrote: > > Hi, > > on 2019/8/13 3:17, Stephen Wilcoxon wrote: > > Perl, on the other hand, can do anything Python can (except stackless) > and, > generally, just as easily. > > > I don't think so specially in AI/ML field. > > Python can handle primitive types much better than ruby/perl can at the > moment. And has much less of an overhead, compared to ruby where almost > everything is a FixNum etc (perl is even worse on types), where Python can > see the actual integer or float and calculations can run at close to C++ > speeds. > > > Python handles types differently than perl6 but certainly not better, even > perl 5 only has a few deficits compared to python mostly it's just > different > > > -- > > .~. In my life God comes first.... > /V\ but Linux is pretty high after that :-D > /( )\ Francis (Grizzly) Smit > ^^-^^ http://www.smit.id.au/ > >
