29
submitted 4 hours ago by Beep@lemmus.org to c/technology@lemmy.world
top 13 comments
sorted by: hot top new old
[-] Awajuk@piefed.social 5 points 3 hours ago

Nice but lacks polish support :(

[-] ExcessShiv@lemmy.dbzer0.com 4 points 2 hours ago* (last edited 2 hours ago)
[-] Awajuk@piefed.social 4 points 2 hours ago

like ๐Ÿ‡ต๐Ÿ‡ฑ polish ๐Ÿ˜… #poland

[-] ExcessShiv@lemmy.dbzer0.com 1 points 2 hours ago

Aah...Polish then, not polish

[-] Awajuk@piefed.social 5 points 1 hour ago

Hahaha, you are right in English!
In Polish we do not use capital letters for languages - Polish is just polski.
We use capital letters for countries, cities etc.
But you are right. In English it is different :D

[-] ExcessShiv@lemmy.dbzer0.com 1 points 1 hour ago

Yeah we don't do that in my native language either...but here we are using English

[-] Awajuk@piefed.social 1 points 1 hour ago

But you are right. In English it is different :D

[-] XLE@piefed.social 3 points 2 hours ago

How much does it cost to run a system that supports it, in 2026 hardware prices? 4B is not the biggest AI model number, but RAM and GPU prices are very daunting thanks to the company releasing this closed-source model

[-] Tetsuo@jlai.lu 1 points 2 hours ago

Closed source model? What do you mean?

Isn't this open weights?

[-] XLE@piefed.social 4 points 2 hours ago

"Open weights" just means you can download the blob they output from their sources. So... Closed source, unless they open it.

Their terminology is just tricky marketing. It would be like calling a closed source program "open executable" or something...

[-] Tetsuo@jlai.lu 1 points 6 minutes ago

My understanding was that the weights were the "essence" of the training.

I think it's a bit misleading to present them as "executables".

I agree that open weights is a bit of marketing mumbo jumbo but I wouldn't say they are akin to a closed source binary.

That being said I was just reading an article about LLM sleeper agents and their trigger words... So you can hide stuff on training and it's fairly hard to spot with just the weights.

But again it's not really like a black box executable. And I've seen many great models that successfully builds on top of an open weights model.

In the end I much prefer open weights to the very popular "'''openAI""" that have opaque training and weights...

[-] Tetsuo@jlai.lu 1 points 2 hours ago

What about the mention of the Apache 2 licence?

[-] Peruvian_Skies@sh.itjust.works 3 points 1 hour ago

Open weights under the Apache license. The model weights are open but not the training process.

this post was submitted on 05 Feb 2026
29 points (78.4% liked)

Technology

80478 readers
4320 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related news or articles.
  3. Be excellent to each other!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
  9. Check for duplicates before posting, duplicates may be removed
  10. Accounts 7 days and younger will have their posts automatically removed.

Approved Bots


founded 2 years ago
MODERATORS