Tucano 2 Cool: Better Open Source LLMs for Portuguese
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911483682619392 |
|---|---|
| author | Corrêa, Nicholas Kluge Sen, Aniket Fatimah, Shiza Falk, Sophia Landgraf, Lennard Kastner, Julia Flek, Lucie |
| author_facet | Corrêa, Nicholas Kluge Sen, Aniket Fatimah, Shiza Falk, Sophia Landgraf, Lennard Kastner, Julia Flek, Lucie |
| contents | We present Tucano 2, a fully open suite of large language models (LLMs) with 0.5-3.7 billion parameters, designed to address certain gaps in open-source development for Portuguese LLMs. Following our previous works, we now extend our dataset, GigaVerbo-v2, to a new degree of quality and scale, while also introducing a new synthetic dataset, GigaVerbo-v2 Synth, aimed at filling missing gaps in GigaVerbo-v2, and two post-training datasets, GigaVerbo-v2 SFT and GigaVerbo-v2 Preferences, that allow Portuguese LLMs to be trained in domains like retrieval augmented generation, coding, tool use, chain-of-thought reasoning, and many other domains of interest. Through extensive ablation studies, we design both pretraining and continual pretraining recipes for the Tucano 2 suite (Base, Instruct, and Think), which achieve state-of-the-art performance on several Portuguese-language modeling benchmarks. We also extend and refine the evaluation harness introduced in our earlier work, yielding a comprehensive evaluation suite that provides strong signals across different pretraining, continual pretraining, and post-training regimes. All artifacts associated with Tucano 2 are openly released, including training recipes, logs, and source code, ensuring that our work is reproducible, accessible, and extendable by the broader Portuguese NLP community. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_03543 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Tucano 2 Cool: Better Open Source LLMs for Portuguese Corrêa, Nicholas Kluge Sen, Aniket Fatimah, Shiza Falk, Sophia Landgraf, Lennard Kastner, Julia Flek, Lucie Computation and Language Artificial Intelligence We present Tucano 2, a fully open suite of large language models (LLMs) with 0.5-3.7 billion parameters, designed to address certain gaps in open-source development for Portuguese LLMs. Following our previous works, we now extend our dataset, GigaVerbo-v2, to a new degree of quality and scale, while also introducing a new synthetic dataset, GigaVerbo-v2 Synth, aimed at filling missing gaps in GigaVerbo-v2, and two post-training datasets, GigaVerbo-v2 SFT and GigaVerbo-v2 Preferences, that allow Portuguese LLMs to be trained in domains like retrieval augmented generation, coding, tool use, chain-of-thought reasoning, and many other domains of interest. Through extensive ablation studies, we design both pretraining and continual pretraining recipes for the Tucano 2 suite (Base, Instruct, and Think), which achieve state-of-the-art performance on several Portuguese-language modeling benchmarks. We also extend and refine the evaluation harness introduced in our earlier work, yielding a comprehensive evaluation suite that provides strong signals across different pretraining, continual pretraining, and post-training regimes. All artifacts associated with Tucano 2 are openly released, including training recipes, logs, and source code, ensuring that our work is reproducible, accessible, and extendable by the broader Portuguese NLP community. |
| title | Tucano 2 Cool: Better Open Source LLMs for Portuguese |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2603.03543 |