language:
- ca
license: apache-2.0
tags:
- catalan
- qa
datasets:
- projecte-aina/catalanqa
- projecte-aina/xquad-ca
model-index:
- name: roberta-base-ca-v2-cased-qa
results:
- task:
type: question-answering
dataset:
type: projecte-aina/catalanqa
name: CatalanQA
metrics:
- name: F1
type: f1
value: 0.895
- task:
type: question-answering
dataset:
type: projecte-aina/xquad-ca
name: XQuAD-Ca
metrics:
- name: F1
type: f1
value: 0.7364
metrics:
- f1
- exact match
widget:
- text: Quan va començar el Super3?
context: >-
El Super3 o Club Super3 és un univers infantil català creat a partir d'un
programa emès per Televisió de Catalunya des del 1991. Està format per un
canal de televisió, la revista Súpers!, la Festa dels Súpers i un club que
té un milió i mig de socis.
- text: Quants eren els germans Marx?
context: >-
Els germans Marx van ser un grup de còmics dels Estats Units que
originàriament estava compost per cinc germans (entre parèntesis els noms
artístics): Leonard (Chico), Adolph (Harpo), Julius (Groucho), Milton
(Gummo) i Herbert (Zeppo).
- text: On van ser els Jocs Olímpics de 1992?
context: >-
Els Jocs Olímpics d'estiu de 1992, oficialment Jocs Olímpics de la XXV
Olimpíada, es van celebrar a la ciutat de Barcelona entre els dies 25 de
juliol i 9 d'agost de 1992. Hi participaren 9.356 atletes (6.652 homes i
2.704 dones) de 169 comitès nacionals, que competiren en 32 esports i 286
especialitats.
- text: Qui va dissenyar la Sagrada Família?
context: >-
El Temple Expiatori de la Sagrada Família, conegut habitualment com la
Sagrada Família, és una basílica catòlica situada a la ciutat de
Barcelona. És un dels exemples més coneguts del modernisme català i un
edifici únic al món, que ha esdevingut tot un símbol de la ciutat. Obra
inacabada de l'arquitecte català Antoni Gaudí, és al barri de la Sagrada
Família, al districte de l'Eixample de la ciutat.
- text: Quin és el tercer volcà més gran de la Terra?
context: >-
El Teide (o Pic del Teide) és un estratovolcà i muntanya de Tenerife,
Illes Canàries (28.27 N, 16.6 O). Amb una altitud de 3718 m sobre el
nivell del mar i amb aproximadament uns 7000 m sobre el llit marí
adjacent, és la muntanya més alta d'Espanya, la muntanya més alta de totes
les illes atlàntiques i el tercer volcà més gran de la Terra.
Catalan BERTa-v2 (roberta-base-ca-v2) finetuned for Question Answering.
Table of Contents
Click to expand
Model description
The roberta-base-ca-v2-cased-qa is a Question Answering (QA) model for the Catalan language fine-tuned from the roberta-base-ca-v2 model, a RoBERTa base model pre-trained on a medium-size corpus collected from publicly available corpora and crawlers (check the roberta-base-ca-v2 model card for more details).
Intended uses and limitations
roberta-base-ca-v2-cased-qa model can be used for extractive question answering. The model is limited by its training dataset and may not generalize well for all use cases.
How to Use
Here is how to use this model:
from transformers import pipeline
nlp = pipeline("question-answering", model="projecte-aina/roberta-base-ca-v2-cased-qa")
text = "Quan va començar el Super3?"
context = "El Super3 o Club Super3 és un univers infantil català creat a partir d'un programa emès per Televisió de Catalunya des del 1991. Està format per un canal de televisió, la revista Súpers!, la Festa dels Súpers i un club que té un milió i mig de socis."
qa_results = nlp(text, context)
print(qa_results)
Limitations and bias
At the time of submission, no measures have been taken to estimate the bias embedded in the model. However, we are well aware that our models may be biased since the corpora have been collected using crawling techniques on multiple web sources. We intend to conduct research in these areas in the future, and if completed, this model card will be updated.
Training
Training data
We used the QA dataset in Catalan called CatalanQA for training and evaluation, and the XQuAD-ca test set for evaluation.
Training procedure
The model was trained with a batch size of 16 and a learning rate of 5e-5 for 5 epochs. We then selected the best checkpoint using the downstream task metric in the corresponding development set and then evaluated it on the test set.
Evaluation
Variable and metrics
This model was finetuned maximizing F1 score.
Evaluation results
We evaluated the roberta-base-ca-v2-cased-qa on the CatalanQA and XQuAD-ca test sets against standard multilingual and monolingual baselines:
| Model | CatalanQA (F1/EM) | XQuAD-Ca (F1/EM) |
|---|---|---|
| roberta-base-ca-v2-cased-qa | 89.50/76.63 | 73.64/55.42 |
| roberta-base-ca-cased-qa | 89.17/77.14 | 69.20/51.47 |
| mBERT | 86.90/74.19 | 68.79/50.80 |
| XLM-RoBERTa | 88.17/75.93 | 72.55/54.16 |
For more details, check the fine-tuning and evaluation scripts in the official GitHub repository.
Additional information
Author
Text Mining Unit (TeMU) at the Barcelona Supercomputing Center ([email protected])
Contact information
For further information, send an email to [email protected]
Copyright
Copyright (c) 2022 Text Mining Unit at Barcelona Supercomputing Center
Licensing information
Funding
This work was funded by the Departament de la Vicepresidència i de Polítiques Digitals i Territori de la Generalitat de Catalunya within the framework of Projecte AINA.
Citation Information
If you use any of these resources (datasets or models) in your work, please cite our latest paper:
@inproceedings{armengol-estape-etal-2021-multilingual,
title = "Are Multilingual Models the Best Choice for Moderately Under-resourced Languages? {A} Comprehensive Assessment for {C}atalan",
author = "Armengol-Estap{\'e}, Jordi and
Carrino, Casimiro Pio and
Rodriguez-Penagos, Carlos and
de Gibert Bonet, Ona and
Armentano-Oller, Carme and
Gonzalez-Agirre, Aitor and
Melero, Maite and
Villegas, Marta",
booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021",
month = aug,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.findings-acl.437",
doi = "10.18653/v1/2021.findings-acl.437",
pages = "4933--4946",
}
Disclaimer
Click to expand
The models published in this repository are intended for a generalist purpose and are available to third parties. These models may have bias and/or any other undesirable distortions.
When third parties, deploy or provide systems and/or services to other parties using any of these models (or using systems based on these models) or become users of the models, they should note that it is their responsibility to mitigate the risks arising from their use and, in any event, to comply with applicable regulations, including regulations regarding the use of Artificial Intelligence.
In no event shall the owner and creator of the models (BSC – Barcelona Supercomputing Center) be liable for any results arising from the use made by third parties of these models.