Multilingual Evaluation Suite supporting 21 European Languages
EuroLingua-GPT
community
AI & ML interests
None defined yet.
Recent Activity
View all activity
Organization Card
EuroLingua-GPT
🧠 What is EuroLingua-GPT?
EuroLingua-GPT is a multilingual large language model initiative led by Fraunhofer IAIS, AI Sweden, and TU Dresden. It aims to build a state-of-the-art open-source LLM tailored for Europe, covering 37 European languages and beyond.
🎯 Project Goal
- Develop a high-performing multilingual LLM optimized for European languages.
- Collect, curate, and evaluate large-scale multilingual datasets.
- Train and align the model using the latest in transformer and instruction-tuning techniques.
- Openly release the model to support research, innovation, and responsible AI development in Europe.
- Training Framework: GitHub - Modalities
🗓️ Project Timeline May 1, 2024 – October 1, 2025
models 0
None public yet
datasets 18
Eurolingua/HPLT3_DE_0.8_Quantile
Viewer
• Updated
• 14.8M • 8
Eurolingua/HPLT3_DE_0.8_Quantile_Adult_Filtered
Updated
• 3
Eurolingua/HPLT3_DE_0.9_Quantile
Viewer
• Updated
• 9.88M • 8
Eurolingua/HPLT3_DE_0.9_Quantile_DiverseQA
Updated
• 3
Eurolingua/test
Viewer
• Updated
• 318k • 6
Eurolingua/HPLT3_DE_0.9_Quantile_Adult_Filtered
Viewer
• Updated
• 9.99M • 17 • 1
Eurolingua/hplt3_edu_scores
Viewer
• Updated
• 1.55B • 734
Eurolingua/hplt3_domains
Viewer
• Updated
• 7.12B • 308
Eurolingua/DCLM-200-100k-exact-dedup
Viewer
• Updated
• 17.7M • 247
Eurolingua/DCLM-200-100k-unfiltered
Viewer
• Updated
• 18.9M • 158 • 1