Hugging Face's logo

LangTech Dev

non-profit
https://www.bsc.es/discover-bsc/organisation/research-departments/language-technologies-unit
Activity Feed Request to join this org

AI & ML interests

This organization is for internal development and testing of models by the Langtech group at BSC. Things published here are not generally useful to people outside of the development team.

Recent Activity

acumplido  updated a Space 4 days ago
langtech-dev/MED-RAG
crodri  updated a Space 10 days ago
langtech-dev/MED-RAG
crodri  published a Space 10 days ago
langtech-dev/MED-RAG
View all activity

langtech-dev's activity

crodri 
posted an update about 4 hours ago
view post
Post
220
At the Language Technologies Unit of the Barcelona Supercomputing Center, we are developing State of the Art Large Language and Voice Models through various national and international projects. It si an exciting time to be working in generative AI!
We are looking for bright and motivated individuals to help us achieve ambitious goals. Our latest opening for the Innovation group that develops powerful and socially useful applications for AI technology might be for you. Check it out here:
https://www.bsc.es/join-us/job-opportunities/3025lsltre2
  • 1 reply
·
acumplido 
updated a Space 4 days ago
crodri 
updated a Space 10 days ago
crodri 
published a Space 10 days ago
MarinaPlius 
updated a Space 14 days ago
jgrivolla 
updated a model about 1 month ago
ankush13r 
updated a model about 2 months ago
jgrivolla 
updated 4 models 4 months ago
crodri 
posted an update 9 months ago
view post
Post
2214
Multilingual RAG optimized models and datasets available from the Language Technologies Unit @ the Barcelona Supercomputing Unit
We are releasing new RAG-optimized multilingual models and dataset, within the AINA project contributions:

projecte-aina/FlorRAG , based on out Bloom Flor6.3b model, capable of RAG in Catalan, Spanish and English

projecte-aina/RAG_Multilingual , a 56K+ instructional dataset with human-like answers created from kernel-of-truth of extractive datasets using a Mixtral8x7b model