www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - ML With Dlang???

reply Ahamed <ahamedrashid.me gmail.com> writes:
.. Soo..  is it possible to learn / make ml model in dlang? 
anyone had made it? or has any clear knowledge about the process? 
( any lib/framework suggestion/ ongoing tools) help me with this! 
THANKS
Aug 01
next sibling parent Sergey <kornburn yandex.ru> writes:
On Saturday, 1 August 2026 at 20:06:51 UTC, Ahamed wrote:
 .. Soo..  is it possible to learn / make ml model in dlang? 
 anyone had made it? or has any clear knowledge about the 
 process? ( any lib/framework suggestion/ ongoing tools) help me 
 with this! THANKS
Yes, its possible. I'm working on several libs/frameworks. But I wouldn't recommend it for newcomers. Better to stick with popular solutions from R/Python. But which kind of models are interesting for you?
Aug 01
prev sibling next sibling parent monkyyy <crazymonkyyy gmail.com> writes:
On Saturday, 1 August 2026 at 20:06:51 UTC, Ahamed wrote:
 .. Soo..  is it possible to learn / make ml model in dlang? 
 anyone had made it? or has any clear knowledge about the 
 process? ( any lib/framework suggestion/ ongoing tools) help me 
 with this! THANKS
you could ask an ai on low settings to make k-means in dlang Theres a few (cpu?) nn's I been meaning to write some custom stuff Id probaly see yes for any small scale stuff, competitive super computer nn's not really go do that with the magic c++
Aug 01
prev sibling next sibling parent Denis F <denis.feklushkin gmail.com> writes:
On Saturday, 1 August 2026 at 20:06:51 UTC, Ahamed wrote:
 .. Soo..  is it possible to learn / make ml model in dlang? 
 anyone had made it? or has any clear knowledge about the 
 process? ( any lib/framework suggestion/ ongoing tools) help me 
 with this! THANKS
Six months ago, there was a lot of hype around Andrej Karpathy's microgpt.py: https://forum.dlang.org/thread/rbfrsdnxjlikmmdeqmgs forum.dlang.org
Aug 03
prev sibling next sibling parent reply Lance Bachmeier <no spam.net> writes:
On Saturday, 1 August 2026 at 20:06:51 UTC, Ahamed wrote:
 .. Soo..  is it possible to learn / make ml model in dlang? 
 anyone had made it? or has any clear knowledge about the 
 process? ( any lib/framework suggestion/ ongoing tools) help me 
 with this! THANKS
Your inquiry isn't precise, so I'll give an answer based on strong assumptions. I'd use either version of this book if the goal is to learn ML from the beginning: https://www.statlearning.com/ I would rewrite or extend the libraries in D. If you want AI assistance converting code and/or math to D, I recommend using Laguna S 2.1, as it's dirt cheap and very capable with D (same could be said for GPT Luna, but there's still a significant price gap compared to Laguna). My answer reflects a view that there's no reason to rewrite working code if you can trivially call it (and that's been a foundation of D from the beginning). What has value is if you're writing new functionality in D.
Aug 03
parent Serg Gini <kornburn yandex.ru> writes:
On Monday, 3 August 2026 at 14:05:02 UTC, Lance Bachmeier wrote:
 My answer reflects a view that there's no reason to rewrite 
 working code if you can trivially call it (and that's been a 
 foundation of D from the beginning). What has value is if 
 you're writing new functionality in D.
Rewriting could be useful for learning though.
Aug 03
prev sibling next sibling parent Meta <jared771 gmail.com> writes:
On Saturday, 1 August 2026 at 20:06:51 UTC, Ahamed wrote:
 .. Soo..  is it possible to learn / make ml model in dlang? 
 anyone had made it? or has any clear knowledge about the 
 process? ( any lib/framework suggestion/ ongoing tools) help me 
 with this! THANKS
https://blog.dlang.org/2026/06/07/teaching-an-ai-to-know-itself-building-a-local-llm-agent-in-d/
Aug 03
prev sibling parent reply Vladimir Panteleev <thecybershadow.lists gmail.com> writes:
On Saturday, 1 August 2026 at 20:06:51 UTC, Ahamed wrote:
 .. Soo..  is it possible to learn / make ml model in dlang? 
 anyone had made it? or has any clear knowledge about the 
 process? ( any lib/framework suggestion/ ongoing tools) help me 
 with this! THANKS
Yes, there are a few libraries. All abandoned, I think. - Most famous, Netflix VectorFlow: https://github.com/Netflix/vectorflow - grain (has some CUDA support): https://github.com/ShigekiKarita/grain - Drain (my own, focuses on design / code clarity rather than features): https://github.com/CyberShadow/Drain
Aug 04
parent reply Serg Gini <kornburn yandex.ru> writes:
On Tuesday, 4 August 2026 at 12:50:35 UTC, Vladimir Panteleev 
wrote:
 Yes, there are a few libraries. All abandoned, I think.
Though these all are DL libs not ML :)
Aug 04
parent reply Vladimir Panteleev <thecybershadow.lists gmail.com> writes:
On Tuesday, 4 August 2026 at 13:57:46 UTC, Serg Gini wrote:
 On Tuesday, 4 August 2026 at 12:50:35 UTC, Vladimir Panteleev 
 wrote:
 Yes, there are a few libraries. All abandoned, I think.
Though these all are DL libs not ML :)
I don't know what definition you're using, but "deep learning" is generally considered to be a kind of machine learning. https://en.wikipedia.org/wiki/Deep_learning https://en.wikipedia.org/wiki/Machine_learning
Aug 04
parent Serg Gini <kornburn yandex.ru> writes:
On Tuesday, 4 August 2026 at 14:27:48 UTC, Vladimir Panteleev 
wrote:
 On Tuesday, 4 August 2026 at 13:57:46 UTC, Serg Gini wrote:
 On Tuesday, 4 August 2026 at 12:50:35 UTC, Vladimir Panteleev 
 wrote:
 Yes, there are a few libraries. All abandoned, I think.
Though these all are DL libs not ML :)
I don't know what definition you're using, but "deep learning" is generally considered to be a kind of machine learning. https://en.wikipedia.org/wiki/Deep_learning https://en.wikipedia.org/wiki/Machine_learning
exactly. DL is a part of ML But many ML algos have different simpler approaches.
Aug 04