Basque Center for Language Technology, a reference center whose aim is to promote research, training, technological transfer and innovation in
#nlproc
&
#voice
.
We are happy to present the Latxa family of open language models, which includes the largest and best model for Basque. It is based on
@Meta
's Llama models ranging between 7-70 billion parameters, with the open license Llama-2.
Nos alegramos de presentar la familia de modelos de lenguaje abiertos Latxa, que tiene el mayor y mejor modelo lingüístico para el euskera. Está basado en los modelos Llama de
@Meta
y contiene modelos de entre 7-70 mil millones de parámetros, todos con licencia abierta Llama-2.
Pozarren aukezten dugu Latxa eredu irekien familia, euskarazko hizkuntza eredurik handiena eta hoberena duena.
@Meta
-ren Llama ereduetan oinarritutakoa eta 7-70 mila miloi parametro arteko ereduak biltzen ditu, Llama-2 delako lizentzia irekia dute.
1/n
Arranca el proyecto de tecnologías lingüísticas del Estado dentro del PERTE Nueva Economía de la Lengua. El proyecto más importante de la historia del Procesado de Lenguaje Natural en España
#PERTE_DeLaLengua
@BSC_CNS
@projecte_aina
We are happy to announce that the inauguration of HiTZ (Basque Center for Language Technology) will be on Thursday November 5th 2020.
Stay tuned for more information!
#nlproc
We are hiring! The center has received funding for several ambitious projects, and is offering five PhD positions and six positions for researchers, both pre- and post-doctoral. Please check 👇🏼
Gaur,
#otsailak11
, emakume eta neska zientzialarien nazioarteko eguna da. Aurten HiTZ zentrotik hizkuntzaren prozesamenduan eta adimen artifizialean lanean ari diren emakume hizkuntzalarien lana azpimarratu nahi dugu.
#emakumeakzientzian
Farewell
#acl2022nlp
! It was a very nice conference. We have presented 2 main conference and 2 workshop papers, and organised 2 workshops.
Really looking forward to go to ACL 2023 in Toronto.
Today
@aormazabalo
has defended his PhD thesis "Towards general attribute controllability in NLP models", supervised by
@artetxem
and
@eagirre
.
Congratulations to the new doctor! 🎉🎉
Director of DARIAH-EU Sally Chambers
@schambers3
is presenting the infrastructure and giving an insight on all the innitiatives and opportunities of the network in the III INTELE workshop.
@DARIAHeu
#INTELE
Hello world! This is the official account of the HiTZ zentroa, the Basque Center for Language Technology, a reference center whose aim is to promote research, training, technological transfer and innovation in this area.
Gaur, otsailak 21,
#amaHizkuntzarenEguna
da. Gure taldean 6 ama hizkuntza baino gehiago ditugu. Guztion artean gure ama hizkuntzetako hitz gogokoenak hautatu ditugu eta irudi hau osatu dugu, zeintzuk dira zuen hitz gogokoenak zuen ama hizkuntzan?
We are at
#METAFORUM2022
.
Today, German Rigau talks at 9:40 about the National LT/AI Strategies in Europe, presenting the work carried out in the ELE
@EuroLangTech
project.
You can follow the streaming here:
#languagetechnologies
Our center members
@oierldl
and
@osainz59
enjoyed the demo day organized by the BETTER program (funded by the US gov.) in Washington, D.C.
Our team (lead by
@bnmin
) presented two demos on few shot information extraction and retrieval tasks.
Atzo Kubako sari akademiko gorena eman zioten Xabier Artola Ixakideari.
20 urteko lankidetza zientifikoarengatik eman diote saria.
Hortik atera da Diccionario Básico Escolar Cubano hiztegia. 300.000 ale eta 1.100 CD saldu dira eskoletan.
ZORIONAK
[EN] 5th
#intele
Webinar
"Distant Reading for European Literary History" (EN ad ES) by Christof Schöch, Borja Navarro and Rosario Arias
📅June 11, 14:00 CET
Registration:
#DigitalHumanities
#DH
We are happy to announce the next webinar!
Speaker:
@VeredShwartz
(The University of British Columbia-Vancouver)
Title: Incorporating Commonsense Reasoning into NLP Models
📅November 3, 2022, 15:30 CET
Registration:
#nlproc
We are glad to announce the keynote talk for the
#TDLE2022
workshop
@lrec2022
, co-organised with
@EuroLangTech
:
"Measuring Cultural Representativeness and Rethinking LT4Al" by Antonios Anastasopoulos from the George Mason University
Atzo, gure taldekide
@eagirre
-k eta
@artetxem
-ek Informatikako
@scie_inform
-
@FundacionBBVA
sariak jaso zituzten: Enekok Aritmel saria eta Mikelek ikertzaile gazteena.
Zorionak! 🎉🎉
Informazio gehiago (es):
📢We are happy to announce an additional webinar in the Language Technology webinar series!
Speaker: Mikel Artetxe
@artetxem
(FAIR, Meta AI)
Title: Is scale all you need?
Date: Jun 24, 2022, 10:00 CET
Registration (free):
Do not miss it!
#nlproc
HiTZ zentroak ere korrikaren alde egin zuen, gure zuzendariak eta zentroko hainbat kide eta kide-ohiek lekukoa eraman zuten larunbatean ilunabarrean.
#korrika2024
#korrika24
Iria de-Dios-Flores is presentig the
@proxectoNos
, Opening routes for the Galician language in the field of language technologies
#TDLE2022
workshop at
#lrec2022
La semana pasada participamos en la primera reunión de la red estratégica CLARIAH-ES en
Alicante, para diseñar la participación en las infraestructuras de investigación europeas CLARIN y DARIAH.
Para más información:
We are happy to announce that our deputy director, German Rigau, will be the pro-term President (2022-2023) of the Global WordNet Association (GWA). Congratulations, German! 🎉🎉
Agur
#acl2022nlp
!
Oso kongresu atsegina izan da. Lau artikulu aurkeztu ditugu, bi konferentzia nagusian eta beste bi workshopetan. Horretaz gain, bi workshop antolatu ditugu.
Hurrengo urtean Toronto ikusiko gara!
Adimen artifizialek albora ez gaitzaten ari gara lanean
@Hitz_Zentroan
. Adimen artifiziala feminista izango da edo ez da izango.
#Martxoak8
Dall·e erabiliz sortutako irudia:
Nuestro subdirector German Rigau ha participado en la mesa redonda sobre Inteligencia Artificial y Procesamiento del Lenguaje Natural aplicado a la generación de conocimiento en el
#CongresoDialnetGlobal
Continuamos con una mesa redonda en el
#CongresoDialnetGlobal
Inteligencia Artificial y Procesamiento del Lenguaje Natural aplicado a la generación de conocimiento con:
Eduardo Bergasa Balda
@ebergasa
Fundación Dialnet
@dialnet
German Rigau, Coordinador Nacional CLARIN y DARIAH
Prest ditugu jada gure 12 A100 de 80Gb VRAM GPU berriak!
Hizkuntzara bideratutako adimen artifiziala egiteko erabiliko ditugu.
Makina hauekin, estatuan aurrenekoz, testu eleaniztunak sortuko eta ulertuko dituzten hizkuntza-eredu erraldoiak sortuko ditugu.
And, yes, there is a seventh member in the picture. We are very excited to announce that
@artetxem
is now officially a part-time academic research collaborator of HiTZ. Ongi etorri Mikel!
Last week we took part on a technical meeting of
@SignONEU
project at
@UPFBarcelona
. It was great to meet in person again and very useful for advancing in the final steps of this project. Thanks to the organizers and other participants!
Los miembros de HiTZ German Rigau y
@iruskieta_m
colaborando con el nodo de CLARIAH-ES IATEXT-ULPGC en Las Palmas de Gran Canaria. Hemos presentado CLARIAH-ES, DARIAH y CLARIN, infraestructuras europeas para las Artes, Humanidades y Ciencias Sociales
Come to our poster!
Izaskun Aldezabal is presenting TZOS, an online terminology database aimed at working on
#Basque
academic
#terminology
collaboratively at
#LREC2022
Spain is officially a
@CLARINERIC
member. In this Social Sciences and Humanities infrastructure HiTZ will mainly provide the resources for Basque and its NLP expertise.
🗣️Ya tenemos fecha para el próximo evento organizado por
#DiverTLes
💜
Mesa Redonda: Impacto de los LLMs en escenarios reales
Se celebrará virtualmente el 🗓️7 de marzo a las ⏱️11.30
Inscripción gratuita requerida
#WomenInNLP
#MujeresEnPLN
#ImpactoPLN
👇
Y el premio a la mejor tesis es para...
And the award to the best thesis goes to...
"Unsupervised Machine Translation"
Zorionak Mikel Artetxe
@artetxem
! 🎉🎉🥳🥳🏆🏆
#SEPLN2021
Levar o
#galego
á sociedade e a economía do S. XXI:
@Xunta
e
@UniversidadeUSC
celebran este luns a primeira xornada de difusión pública do
#ProxectoN
ós, o maior investimento público da historia en favor da lingua propia, desenvolvido por CiTIUS e
@ILG_USC
#LRECCOLING2024
kongresuan izango gara hurrengo artikuluekin, horietariko batzuk munduko beste ikertzaileekin batera egindakoak
We will be at
#LRECCOLING2024
with the following papers, some of them co-authored with collegues around the world
#Tecnolox
íasDaLinguaxe udako ikastaroan gaude. Oraintxe hasiko da
@glabaka
itzulpen automatikoari buruz eta bihar Inma Hernaezen eta German Rigauren txanda izango da.
@proxectoNos
We are happy to announce that we are hosting the 12th International Global Wordnet Conference
#GWC2023
in Donostia/San Sebastian.
The conference will be in January 23-27, 2023.
#METAFORUM2022
-en gaude!
Gaur, 9:40ean German Rigauk hizkuntza teknologiei eta adimen artifiziari buruzko estatuen plangintzak izango ditu hizpide (ELE
@EuroLangTech
proiektuan egindako lana)
Saioa online jarrai daiteke:
#hizkuntzateknologiak
Yesterday German Rigau presented and chaired a session on National Language Technology and AI Programmes in European
#METAFORUM
2023. HiTZ Center is actively taking part in making Basque and Spanish fully digital languages.
Tomorrow,
@eagirre
gives the invited talk "Few-shot Information Extraction is Here: Pre-train, Prompt and Entail" at
#sigir2022
Bihar
@eagirre
-k "Few-shot Information Extraction is Here: Pre-train, Prompt and Entail" hitzaldi gonbidatua emango du.
HiTZ webinars come back!
#nlproc
We are happy to announce that our next speaker will be Hinrich Schütze (
@LMU_Muenchen
)
Title: Humans Learn From Task Descriptions and So Should Our Models
📅 4 Nov, 15:00 CET
Registration (free) form:
BIDEOA |▶️ HiTZ Zentroak euskararentzako hizkuntza eredurik handiena garatu du:
#Latxa
Aitor Soroa,
@Hitz_zentroa
: "Eredu ingelesa hartu genuen eta euskara irakatsi genion"
👉
#GureanGaur
@upvehu
92 investigadores e investigadoras de la UPV/EHU, entre los más influyentes del mundo. El ranking de la Universidad de Stanford incluye el grupo del 2 % más citado. 👏👏👏Zorionak!!!
#Investigaci
ón
#Universidad
➡️
En
@elpais_tec
, nuestro subdirector German Rigau afirma sobre el modelo de lenguaje que anunció
@desdelamoncloa
@sanchezcastejon
que solo unos pocos grupos son capaces de programar un modelo de este tipo y que todos están en centros de investigación o universidades.
Congrats to the new class of ACL fellows, with groundbreaking work in parsing, MT, discourse, grounded NLP, WSD, dialog, multilingual NLP:
Joakim Nivre, Qun Liu, Luke Zettlemoyer, Anette Frank, Joyce Chai, Eneko Agirre, Chengqing Zong, Amanda Stent
👏🎉
"Handwritten Text Recognition and Basque exam papers", a presentation by
@iruskieta_m
and
@ekainarrieta
at the IMPACT 10th Anniversary Workshop and Writing Sprint hosted at University of Alicante on 12-14 December, 2022.
Today we can finally present our last work on LLMs for personalized Information Extraction! We named it GoLLIE, a LLM able to follow detailed annotation guidelines to perform IE.
📒Blog:
📖Paper:
🐶GitHub:
Gaur,
#Otsailak11
, Hitz zentroak Emakume eta Neska Zientzialarien Nazioarteko Egunarekin bat egiten du, baina ez ahaztu
#emakumeakzientzian
egunero kontua dela!!!
#Zientzia
BADA nesken kontua!