--- license: apache-2.0 language: - ca - es - en base_model: gplsi/Aitana-7B-S-base-1.0 tags: - valencian - spanish - english - text-generation - instruct - alia - gplsi datasets: - OpenAssistant/oasst2 - OpenAssistant/oasst1 - BSC-LT/m-personas - projecte-aina/RAG_Multilingual - facebook/flores - CohereLabs/aya_dataset - Unbabel/TowerBlocks-v0.2 - projecte-aina/MentorES - databricks/databricks-dolly-15k - tatsu-lab/alpaca - openai/gsm8k - Open-Orca/OpenOrca - HuggingFaceH4/no_robots - projecte-aina/CoQCat - gplsi/boua_parallel - allenai/scifact - somosnlp/LingComp_QA - somosnlp/instruct-legal-refugiados-es library_name: transformers pipeline_tag: text-generation --- # Aitana-7B-S-Instruct-v0.1 **Aitana-7B-S-Instruct-v0.1** is an instruction-tuned generative language model from the **Aitana family**, developed by the [GPLSI (Language and Information Systems Group)](https://gplsi.dlsi.ua.es/) at the University of Alicante. Built on [gplsi/Aitana-7B-S-base-1.0](https://huggingface.co/gplsi/Aitana-7B-S-base-1.0), this model has been fine-tuned to follow instructions effectively across Valencian, Spanish, and English, with particular emphasis on enhancing Valencian language capabilities. ## Table of Contents - [Model Description](#model-description) - [Evaluation](#evaluation) - [Training Data](#training-data) - [Intended Uses](#intended-uses) - [How to Use](#how-to-use) - [Additional Information](#additional-information) ## Model Description | Property | Value | |----------|-------| | **Base Model** | [gplsi/Aitana-7B-S-base-1.0](https://huggingface.co/gplsi/Aitana-7B-S-base-1.0) | | **Architecture** | Transformer decoder-only | | **Parameters** | ~7.77B | | **Languages** | Valencian, Spanish, English | | **License** | Apache 2.0 | Aitana-7B-S-Instruct-v0.1 is an instruction-tuned variant of Aitana-7B-S-base-1.0, fine-tuned on multilingual instruction data to follow user prompts and generate helpful responses across Valencian, Spanish, and English. ## Training Data This model was instruction fine-tuned on the ALIA Instruction/v12 dataset, composed of the following sources: | Dataset ID | Name | Languages | Source | |------------|------|-----------|--------| | ins1 | OpenAssistant2 (OASST2) | CA, EN, ES, VA | [OpenAssistant/oasst2](https://huggingface.co/datasets/OpenAssistant/oasst2) | | ins2 | OpenAssistant1 (OASST1) | CA, VA | [OpenAssistant/oasst1](https://huggingface.co/datasets/OpenAssistant/oasst1) | | ins3 | M-Personas | CA, EN, ES, VA | [BSC-LT/m-personas](https://huggingface.co/datasets/BSC-LT/m-personas) | | ins4 | RAG Multilingual | CA, EN, ES, VA | [projecte-aina/RAG_Multilingual](https://huggingface.co/datasets/projecte-aina/RAG_Multilingual) | | ins5 | FLORES | CA, EN, ES | [facebook/flores](https://huggingface.co/datasets/facebook/flores) | | ins6 | Aya Dataset | EN, ES, VA | [CohereLabs/aya_dataset](https://huggingface.co/datasets/CohereLabs/aya_dataset) | | ins7 | TowerBlocks | EN, ES | [Unbabel/TowerBlocks-v0.2](https://huggingface.co/datasets/Unbabel/TowerBlocks-v0.2) | | ins8 | Mentor / Mentores | CA, ES, VA | [projecte-aina/MentorES](https://huggingface.co/datasets/projecte-aina/MentorES) / [projecte-aina/MentorCA](https://huggingface.co/datasets/projecte-aina/MentorCA) | | ins9 | Dolly / Dolly 3K | CA, EN, VA | [databricks/databricks-dolly-15k](https://huggingface.co/datasets/databricks/databricks-dolly-15k) | | ins10 | Alpaca | EN, VA | [tatsu-lab/alpaca](https://huggingface.co/datasets/tatsu-lab/alpaca) | | ins11 | GSM8K | EN, VA | [openai/gsm8k](https://huggingface.co/datasets/openai/gsm8k) | | ins12 | OpenOrca | EN | [Open-Orca/OpenOrca](https://huggingface.co/datasets/Open-Orca/OpenOrca) | | ins13 | No Robots | EN | [HuggingFaceH4/no_robots](https://huggingface.co/datasets/HuggingFaceH4/no_robots) | | ins14 | CoQCA / CoQCat | CA, VA | [projecte-aina/CoQCat](https://huggingface.co/datasets/projecte-aina/CoQCat) | | ins15 | BOUA | ES | [gplsi/boua_parallel](https://huggingface.co/datasets/gplsi/boua_parallel) | | ins16 | SciFact | VA | [allenai/scifact](https://huggingface.co/datasets/allenai/scifact) | | ins17 | LingComp QA | VA | [somosnlp/LingComp_QA](https://huggingface.co/datasets/somosnlp/LingComp_QA) | | ins18 | Instruct Legal Refugiados | VA | [somosnlp/instruct-legal-refugiados-es](https://huggingface.co/datasets/somosnlp/instruct-legal-refugiados-es) | | ins19 | Amic-Paralelo | ES | — | The model was NOT instruction-tuned on Catalan data, though some Catalan appears in multilingual datasets. ## Intended Uses This model can be used for: - **Instruction following** in Valencian, Spanish, and English - **Chat and conversational applications** requiring multilingual support - **Text generation** with task-specific prompting - **Domain-specific applications** in administrative, legal, or tourism contexts > **Note**: As an instruction-tuned model, it is designed to follow user prompts and generate helpful responses. It is not intended for use as a factual knowledge base. ## How to Use ### Transformers ```python import torch from transformers import pipeline, AutoTokenizer model_id = "gplsi/Aitana-7B-S-Instruct-v0.1" tokenizer = AutoTokenizer.from_pretrained(model_id) generator = pipeline( "text-generation", model=model_id, tokenizer=tokenizer, torch_dtype=torch.bfloat16, device_map="auto", ) # Valencian example text = "Explica què són les Corts Valencianes i quina funció tenen." result = generator(text, do_sample=True, top_k=10, max_new_tokens=100) print(result[0]['generated_text']) # Spanish example text = "Describe las principales funciones del gobierno autonómico valenciano." result = generator(text, do_sample=True, top_k=10, max_new_tokens=100) print(result[0]['generated_text']) # English example text = "Explain the role of tourism in the Valencian Community economy." result = generator(text, do_sample=True, top_k=10, max_new_tokens=100) print(result[0]['generated_text']) ``` ## Evaluation In the following table, we present the results obtained with different benchmarks from [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) in comparison with [Salamandra-7B-Instruct](https://huggingface.co/BSC-LT/Salamandra-7B-Instruct). The results reflect the instruction-tuned performance of both models. ### Normalized score per language | Language |Salamandra-7B-Instruct| Aitana-7B-S-Instruct-v0.1 | |----------|----------|----------| | Spanish | **0.236** | 0.219 | | Catalan | **0.343**| 0.304 | | English | 0.300 | **0.303** | | Valencian | 0.546 | **0.600** | ### Valencian #### Classification Benchmarks | Dataset | Lang. | Task | Metric | Salamandra-7B-Instruct | Aitana-7B-S-Instruct-v0.1 | |------------------------------|--------|----------------------------|-------------|---------------|-----------------------| | XNLI | va |Natural Language Inference | acc | **0.552** | 0.534 | #### Generation Benchmarks | Dataset | Lang. | Task | Metric | Salamandra-7B-Instruct | Aitana-7B-S-Instruct-v0.1 | |------------------------------|--------|----------------------------|-------------|---------------|-----------------------| | Cocoteros | va |Reading Comprehension | bleu | 6.391 | **8.929** | | Phrases ca-va | va-ca |Translation - Adaptation | bleu | 67.980 | **81.743** | | Phrases va-ca | va-ca |Translation - Adaptation | bleu | 79.375 | **83.501** | | Phrases va-es | va-es |Translation | bleu | 63.104 | **80.329** | | Phrases es-va | es-va |Translation | bleu | 51.64 | **63.95** | | Truthfulqa_va | va | Truthfulness | bleu_acc | **0.454** | 0.412 | ### Catalan #### Classification Benchmarks | Dataset | Lang. | Task | Metric | Salamandra-7B-Instruct | Aitana-7B-S-Instruct-v0.1 | |------------------------------|--------|---------------------------|-------------|---------------|-----------------------| | Belebele Cat_latn | ca | Reading Comprehension | acc | **0.718** | 0.581 | | COPA | ca | Commonsense Reasoning | acc | **0.824** | 0.822 | | XStoryCloze | ca | Commonsense Reasoning | acc | **0.708** | 0.678 | | OpenBookQA | ca | Question Answering | acc | **0.374** | 0.36 | | PAWS | ca | Paraphrasing | acc | **0.671** | 0.662 | | PiQA | ca | Question Answering | acc | **0.718** | 0.722 | | ARC Easy | ca | Question Answering | acc | 0.686 | **0.713** | | ARC Challenge | ca | Question Answering | acc | 0.425 | **0.435** | | XNLI | ca | Natural Language Inference| acc | **0.559** | 0.540 | | Teca | ca | Natural Language Inference| acc | **0.557** | 0.522 | | WNLI | ca | Natural Language Inference| acc | **0.592** | 0.479 | | Catcola | ca | Linguistic Acceptability | acc | 0.660 | **0.687** | | Catcola | ca | Linguistic Acceptability | mcc | **0.170** | 0.156 | | Catalanqa | ca | Question Answering | F1 | **0.576** | 0.526 | | Mgsm direct | ca | Math | exact match | **0.02** | 0.004 | | Catalanqa | ca | Question Answering | exact match | **0.259** | 0.176 | | Xquad | ca | Question Answering | exact match | **0.228** | 0.157 | | Xquad | ca | Question Answering | F1 | **0.507** | 0.451 | #### Generation Benchmarks | Dataset | Lang. | Task | Metric | Salamandra-7B-Instruct | Aitana-7B-S-Instruct-v0.1 | |------------------------------|--------|--------------------------|--------|----------------|-----------------------| | Cabreu abstractive | ca | Summarization | bleu | 8.60 | **10.10** | | Cabreu extractive | ca | Summarization | bleu | **39.10** | 28.37 | | Cabreu extreme | ca | Summarization | bleu | 3.21 | **3.86** | ### Spanish #### Classification Benchmarks | Dataset | Lang. | Task | Metric | Salamandra-7B-Instruct | Aitana-7B-S-Instruct-v0.1 | |------------------------------|--------|---------------------------|-------------|---------------|-----------------------| | Belebele | es | Reading Comprehension | acc | **0.698** | 0.590 | | PAWS | es | Paraphrasing | acc | **0.629** | 0.626 | | XNLI | es | Natural Language Inference| acc | **0.487** | 0.485 | | WNLI | es | Natural Language Inference| acc | **0.549** | 0.493 | | XStoryCloze | es | Commonsense Reasoning | acc | 0.674 | **0.676** | | Escola | es | Linguistic Acceptability | acc | 0.577 | **0.681** | | Escola | es | Linguistic Acceptability | mcc | **0.179** | 0.178 | | OpenbookQA | es | Question Answering | acc | 0.374 | **0.392** | | MGSM Direct | es | Math | exact match | **0.100** | **0.100** | | XQUAD | es | Question Answering | exact match | **0.189** | 0.087 | | XQUAD | es | Question Answering | F1 | **0.467** | 0.413 | #### Generation Benchmarks | Dataset | Lang. | Task | Metric | Salamandra-7B-Instruct | Aitana-7B-S-Instruct-v0.1 | |------------------------------|--------|---------------------|---------|----------------|-----------------------| | Cocoteros | es |Reading Comprehension| bleu | 6.306 | **8.680** | | XLSum | es | Summarization | bleu | **2.048** | 1.502 | ### English #### Classification Benchmarks | Dataset | Lang. | Task | Metric | Salamandra-7B-Instruct | Aitana-7B-S-Instruct-v0.1 | |------------------------------|--------|----------------------------|-------------|---------------|-----------------------| | Arc Challenge | en | Question Answering | acc | 0.478 | **0.523** | | Arc Easy | en | Question Answering | acc | 0.780 | **0.811** | | Belebele | en | Reading Comprehension | acc | **0.769** | 0.622 | | PAWS | en | Paraphrasing | acc | 0.655 | **0.677** | | XNLI | en | Natural Language Inference | acc | 0.534 | **0.555** | | XStoryCloze | en | Commonsense Reasoning | acc | **0.729** | 0.716 | | OpenBookQA | en | Question Answering | acc | **0.348** | 0.340 | | PiQA | en | Question Answering | acc | 0.781 | **0.784** | | Social iqa | en | Question Answering | acc | 0.520 | **0.524** | | WNLI | en | Natural Language Inference | acc | **0.493** | **0.493** | | MGSM Direct | en | Math | exact match | 0.080 | **0.200** | | TriviaQA | en | Question Answering | exact match | 0.204 | **0.433** | ### Judge Evaluation The model was also evaluated using an LLM-as-judge approach across different task categories. The scores below represent the average rating (1-5 scale, 5 being best) and standard deviation for each task category, comparing Aitana-7B-S-Instruct-v0.1 against Salamandra-7B-Instruct. | Task Category | Salamandra-7B-Instruct | Aitana-7B-S-Instruct-v0.1 | |---------------|------------------------|---------------------------| | CommonSense reasoning | 2.637 / 1.295 | **2.989 / 1.200** | | Maths | 2.386 / 1.536 | **2.584 / 1.474** | | Paraphrasing | 3.725 / 0.967 | **3.927 / 0.981** | | Reading comprehension | **3.472 / 1.015** | 3.420 / 1.268 | | Summarization | **2.369 / 0.932** | 1.862 / 0.713 | | Translation | 3.770 / 0.580 | **3.895 / 0.814** | | **Overall Avg** | 3.060 / 1.054 | **3.113 / 1.075** | ## Additional Information ### Author The model has been developed by the **Language and [Information Systems Group (GPLSI)](https://gplsi.dlsi.ua.es/)** and the **[Centro de Inteligencia Digital (CENID)](https://cenid.es)**, both part of the **[University of Alicante (UA)](https://www.ua.es/es/)**, as part of their ongoing research in **Natural Language Processing (NLP)**. ### Part of the Aitana Family This model is part of the Aitana model family developed by the GPLSI research group, which includes: - [gplsi/Aitana-7B-S-base-1.0](https://huggingface.co/gplsi/Aitana-7B-S-base-1.0) - Base version (1.0) of the 7B model - [gplsi/Aitana-7B-S-Instruct-v0.1](https://huggingface.co/gplsi/Aitana-7B-S-Instruct-v0.1) - Instruction-tuned 7B model - [gplsi/Aitana-2B-S](https://huggingface.co/gplsi/Aitana-2B-S) - Valencian-focused 2B model - [gplsi/Aitana-2B-S-base-1.0](https://huggingface.co/gplsi/Aitana-2B-S-base-1.0) - Base version (1.0) of the 2B model - [gplsi/Aitana-2B-S-Instruct-v0.1](https://huggingface.co/gplsi/Aitana-2B-S-Instruct-v0.1) - Instruction-tuned 2B model - [gplsi/Aitana-2B-S-Instruct-Aligned-v0.1](https://huggingface.co/gplsi/Aitana-2B-S-Instruct-Aligned-v0.1) - DPO-aligned instruction-tuned 2B model - [gplsi/Aitana-6.3B](https://huggingface.co/gplsi/Aitana-6.3B) - Larger 6.3B parameter model - [gplsi/Aitana-TA-2B-S](https://huggingface.co/gplsi/Aitana-TA-2B-S) - Translation model (Spanish ↔ Valencian) - [gplsi/Aitana-2B-S-LF](https://huggingface.co/gplsi/Aitana-2B-S-LF) - 2B Text Generation variant - [gplsi/Aitana-2B-S-tourism-base-1.0](https://huggingface.co/gplsi/Aitana-2B-S-tourism-base-1.0) - Domain-specific base model focused on Tourism - [gplsi/Aitana-tourism-mb-encoder-1.0](https://huggingface.co/gplsi/Aitana-tourism-mb-encoder-1.0) - Tourism domain Fill-Mask/Encoder model - [gplsi/Aitana-FraudDetection-R-1.0](https://huggingface.co/gplsi/Aitana-FraudDetection-R-1.0) - Text Classification model for Fraud Detection ### Funding This work is funded by the **Ministerio para la Transformación Digital y de la Función Pública**, co-financed by the **EU – NextGenerationEU**, within the framework of the project *Desarrollo de Modelos ALIA*. ### Acknowledgments We would like to express our gratitude to all individuals and institutions that have contributed to the development of this work. Special thanks to: - [Language Technologies Laboratory at Barcelona Supercomputing Center](https://www.bsc.es/es/discover-bsc/organisation/research-structure/language-technologies-laboratory) - [Centro Vasco de Tecnología de la Lengua (HiTZ)](https://www.hitz.eus/es) - [Centro Singular de Investigación en Tecnologías Inteligentes (CiTIUS)](https://citius.gal/) - [Sistemas Inteligentes de Acceso a la Información (SINAI)](https://www.ujaen.es/investigacion-y-transferencia/grupos-de-investigacion/sistemas-inteligentes-de-acceso-la-informacion-sinai) - [Instituto Universitario de Investigación Informática (IUII)](https://web.ua.es/es/iuii/) - [Leonardo HPC System](https://leonardo-supercomputer.cineca.eu/) - [European supercomputing ecosystem (EUROHPC)](https://www.eurohpc-ju.europa.eu/) We also acknowledge the financial, technical, and scientific support of the **Ministerio para la Transformación Digital y de la Función Pública - Funded by EU – NextGenerationEU within the framework of the project Desarrollo de Modelos ALIA**, whose contribution has been essential to the completion of this research. ### License [Apache License, Version 2.0](https://www.apache.org/licenses/LICENSE-2.0) ### Disclaimer This model is intended for general purposes and is available under a permissive Apache License 2.0. Be aware that the model may have biases and/or undesirable outputs. Users deploying systems based on this model are responsible for mitigating risks and complying with applicable AI regulations. ### Reference ```bibtex @misc{gplsi-aitana-7B-S-Instruct-v0.1, author = {Sepúlveda-Torres, Robiert and Martínez-Murillo, Iván and Grande, Eduardo and Galiano, Santiago and Consuegra-Ayala, Juan Pablo and Miró Maestre, María and Canal-Esteve, Miquel and Bonora, Mar and Gutierrez, Yoan and Abreu Salas, José Ignacio and Lloret, Elena and Montoyo, Andrés and Muñoz-Guillena, Rafael and Palomar, Manuel}, title = {Aitana 7B Instruct: Instruction-tuned model for Valencian, Spanish and English}, year = {2026}, institution = {Language and Information Systems Group (GPLSI) and Centro de Inteligencia Digital (CENID), University of Alicante (UA)}, howpublished = {\url{https://huggingface.co/gplsi/Aitana-7B-S-Instruct-v0.1}}, note = {Accessed: 2026-05-11} } ``` --- **Copyright © 2026 Language and Information Systems Group (GPLSI) and Centro de Inteligencia Digital (CENID), University of Alicante (UA). Distributed under the Apache License 2.0.**