HPLT

community

https://hplt-project.org/

hplt_eu

hplt-project

Activity Feed Request to join this org

AI & ML interests

Web as a corpus, Large Language Models, Machine Translation, Language Technologies, Natural Language Processing

Recent Activity

mbanon authored a paper 4 days ago

A New Massive Multilingual Dataset for High-Performance Language Technologies

mbanon authored a paper 4 days ago

FastSpell: the LangId Magic Spell

gramirez-prompsit updated a dataset 4 days ago

HPLT/HPLT2.0_cleaned

View all activity

HPLT's activity

mbanon

authored 2 papers 4 days ago

A New Massive Multilingual Dataset for High-Performance Language Technologies

Paper • 2403.14009 • Published Mar 20, 2024 • 1

FastSpell: the LangId Magic Spell

Paper • 2404.08345 • Published Apr 12, 2024

gramirez-prompsit

updated a dataset 4 days ago

HPLT/HPLT2.0_cleaned

Viewer • Updated 4 days ago • 10.6B • 88.7k • 7

oepen

authored a paper 27 days ago

The Impact of Copyrighted Material on Large Language Models: A Norwegian Perspective

Paper • 2412.09460 • Published about 1 month ago • 5

ltgoslo

in HPLT/hplt_bert_base_ja about 1 month ago

Adding `safetensors` variant of this model

#2 opened about 2 months ago by

SFconvertbot

ltgoslo

in HPLT/hplt_bert_base_fr about 1 month ago

Adding `safetensors` variant of this model

#1 opened 2 months ago by

SFconvertbot

ltgoslo

in HPLT/hplt_bert_base_lv about 1 month ago

Adding `safetensors` variant of this model

#1 opened about 2 months ago by

SFconvertbot

ltgoslo

in HPLT/hplt_bert_base_ur about 1 month ago

Adding `safetensors` variant of this model

#1 opened about 1 month ago by

SFconvertbot

BramVanroy

posted an update about 1 month ago

Post

506

In the spirit of "Better late than never", I've finally written a brief overview paper for GEITje 7B Ultra. Initially released 10 months ago (oops), but still reaching around 1300 monthly downloads across the HF ecosystem (not including ollama).

GEITje 7B Ultra: A Conversational Model for Dutch (2412.04092)

While the paper discusses the model a little bit, I especially wanted to write about the datasets, which to this day seem an important asset for Dutch LLM training (SFT and preference tuning). We have a long way to go for Dutch, but publishing transparent and reproducible artefacts seems an important step to me, alongside having open discussions about data, bias, architectures.

In that spirit, thanks are in order for the creation of GEITje 7B Ultra and all related datasets:

- Michiel Buisman and UWV for providing the means to create the datasets
- Flemish Supercomputer Center (VSC) for the compute
- The Hugging Face Fellows and rest of the team for their discussions and insights
- The Dutch NLP community, notably @Rijgersberg for building the base GEITje model and the fruitful discussions we've had

More to come, step by step!

BramVanroy/geitje-7b-ultra-65c1ee010ad80fd1f6a8f208