Sabiá-3 Technical Report

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Abonizio, Hugo, Almeida, Thales Sales, Laitz, Thiago, Junior, Roseval Malaquias, Bonás, Giovana Kerche, Nogueira, Rodrigo, Pires, Ramon
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909560569069568
author Abonizio, Hugo
Almeida, Thales Sales
Laitz, Thiago
Junior, Roseval Malaquias
Bonás, Giovana Kerche
Nogueira, Rodrigo
Pires, Ramon
author_facet Abonizio, Hugo
Almeida, Thales Sales
Laitz, Thiago
Junior, Roseval Malaquias
Bonás, Giovana Kerche
Nogueira, Rodrigo
Pires, Ramon
contents This report presents Sabiá-3, our new flagship language model, and Sabiazinho-3, a more cost-effective sibling. The models were trained on a large brazilian-centric corpus. Evaluations across diverse professional and academic benchmarks show a strong performance on Portuguese and Brazil-related tasks. Sabiá-3 shows large improvements in comparison to our previous best of model, Sabia-2 Medium, especially in reasoning-intensive tasks. Notably, Sabiá-3's average performance matches frontier LLMs, while it is offered at a three to four times lower cost per token, reinforcing the benefits of domain specialization.
format Preprint
id arxiv_https___arxiv_org_abs_2410_12049
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Sabiá-3 Technical Report
Abonizio, Hugo
Almeida, Thales Sales
Laitz, Thiago
Junior, Roseval Malaquias
Bonás, Giovana Kerche
Nogueira, Rodrigo
Pires, Ramon
Computation and Language
Artificial Intelligence
This report presents Sabiá-3, our new flagship language model, and Sabiazinho-3, a more cost-effective sibling. The models were trained on a large brazilian-centric corpus. Evaluations across diverse professional and academic benchmarks show a strong performance on Portuguese and Brazil-related tasks. Sabiá-3 shows large improvements in comparison to our previous best of model, Sabia-2 Medium, especially in reasoning-intensive tasks. Notably, Sabiá-3's average performance matches frontier LLMs, while it is offered at a three to four times lower cost per token, reinforcing the benefits of domain specialization.
title Sabiá-3 Technical Report
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2410.12049