Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

EuroLLM: LLM made in Europe built to support all 24 official EU languages

eurollm.io

Post reply on HN
Print view View original

111–120 of 629 posts

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#111

Post by actionfromafar »

krzyk wrote: post #76
asveikau wrote: post #47

Earlier quoted context omitted.

See the section "historical dispute". I think some people get touchy about them being lumped together if their last period of commonality (per the article) was 1400 BCE. For comparison, I believe all the Slavic languages were mutually intelligible around 1200 AD. But much more recently than this, in the last few centuries, there have been notable attempts by east slavs to absorb the Baltic language cultures and deny…

I doubt that South Slavic and West/East Slavic were mutually intelligible at 1200 AD. I doubt West and East Slavic were. But inside those geographic groups they probably were (Czech and Polish AFAIR were around that time).

Depends on your standards, too. Even today, any pair of slavic speakers should have a head start in understanding each other. Put them next to each other for a month and they should be talking, at least about basic everyday things.

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#112

Post by tensor »

sunaookami wrote: post #85
melvinmelih wrote: post #20

Earlier quoted context omitted.

> because they are trained on multilingual data But they were not trained on government-sanctioned homegrown EU data.

Who in their right mind would use this?

I'd use a model trained on a targeted and curated data set over one trained on all the crap on the internet any day.

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#113

Post by KaiserPro »

whimsicalism wrote: post #104
deaux wrote: post #99

Earlier quoted context omitted.

> What's REALLY much more important though if you want to be a part of the AI race and I've posted for years here with @euaccofficial is to make Europe a really extremely attractive place to start and run an AI business. Remove regulatory obstructions and give tax discounts for startups. Let them build a business first that can compete worldwide and once they make enough money (let's say $100M/y), then slowly start a…

> When you talk to most EU business owners, even in tech, the limiting factor isn't regulations. This being the #1 reason is such a tired trope. Okay, what is the limiting factor? Because when I talk to EU business owners (admittedly, very few) - they point to lack of big EU capital markets, which is directly downstream of the policy environment. And when I talk to top EU human capital, they all point to the lack of…

Money.

Why work in the "europoor" countries when you can go to america and earn megabucks.

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#114

Post by arbuge »

ggsp wrote: post #58
ebb_earl_co wrote: post #34

Earlier quoted context omitted.

What is the name of Maltese in Maltese? Like “el español” in Spanish, it’s neat to know what languages call themselves

Wikipedia says it's "Malti"

Il-Malti to be precise. Il- means "the" and changes its meaning to that of the language. Malti alone would mean a Maltese person.

Source: I'm also Maltese.

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#115

Post by ffsm8 »

whimsicalism wrote: post #91
tokai wrote: post #87

Earlier quoted context omitted.

[flagged]

And your comment is just a totally normal way of engaging with strangers on the internet :)

Tbf, he's kinda on point.

The EU Parlament has it's issues, and it's atrocious how corrupt politicians like e.g. von der Laien fail upwards into leadership roles there, but they're both massively exaggerating the issues to a degree that it's hard to call the stated opinions anything but unhinged.

Its kinda like calling a Chihuahua "apex predator". completely out of touch with reality.

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#116

Post by tensor »

mirekrusin wrote: post #36
lm28469 wrote: post #24

Earlier quoted context omitted.

Meh, it depends a lot on the dataset, which are heavily skewed towards the main languages. For example they almost always confuse Czech and Slovak and often swap one for the other in middle of chats

But the only way to unskew it is to remove main language data because there isn't really any to add, no?

You can also correctly bias your sampling so that when selecting new training instances each language is chosen equally. Generally the diversity of data is good, unless that data is "wrong" which, ironically, is probably most of the internet, but I digress.

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#117

Post by contravariant »

mytailorisrich wrote: post #95
ChrisMarshallNY wrote: post #84

Earlier quoted context omitted.

Flemish? I remember watching a TV show in Flemish ( Hotel Beau Séjour [0]), so it's prevalent enough to invest that kind of money into. What about Basque? Is that too controversial? [0] https://en.wikipedia.org/wiki/Hotel_Beau_Séjour

I think those 24 languages reflect all the languages that are official languages at country level. So for instance, Basque is not an official language of any country (only French in France and Spanish/Castilian in Spain). Belgium's official languages are French, Dutch, and German, "Flemish" is only a local variant of Dutch (Belgian French is also only a local variant of French).

Official is a weird concept though. Turns out Dutch law never really bothered to define an official language, Dutch simply is the de facto standard and is required for a lot of things making it effectively the standard. This makes Dutch Sign Language the only language officially recognised by law. An attempt to recognise Frysian and Dutch as official languages in the constitution failed.

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#118

Post by RobotToaster »

lm28469 wrote: post #24
loandbehold wrote: post #10

Aren't all frontier models already able to use all these languages? Support for specific languages doesn't need to be built in, LLMs support all languages because they are trained on multilingual data.

Meh, it depends a lot on the dataset, which are heavily skewed towards the main languages. For example they almost always confuse Czech and Slovak and often swap one for the other in middle of chats

Aren't they about as different as American English and British English?

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#119

Post by arbuge »

Raed667 wrote: post #40
purrcat259 wrote: post #16

Earlier quoted context omitted.

I read, write and speak Maltese, AMA if you are curious about the language.

Tunisians claim they can understand Maltese with minimum effort, is it reciprocal? How close is Maltese to arabic / tunisian dialect ?

Not sure which Tunisians are claiming this but they'd definitely need a lot more than minimum effort. Maltese split off from Arabic around 1k years ago. The two languages sound pretty different, and are written with different alphabets.

Re: EuroLLM: LLM made in Europe built to support all 24 official EU languages

#120

Post by sublimefire »

kaato137 wrote: post #29
jim180 wrote: post #22

Earlier quoted context omitted.

Lithuanian and Latvian are Baltic languages. Nothing to do with Slavic...

Balto-Slavic branch divides into Baltic and Slavic language groups so nothing wrong here

It is just one of the theories, there is no clear evidence to suggest that Baltic and Slavic were the same language thousands of years ago.
Post reply on HN