datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
mc4-es-sampled50 million documents in Spanish extracted from mC4 applying perplexity sampling via mc4-sampling: "https://huggingface.co/datasets/bertin-project/mc4-sampling". Please, refer to BERTIN Project. The original dataset is the Multlingual Colossal, Cleaned version of Common Crawl's web crawl corpus (mC4), based on the Common Crawl dataset: "https://commoncrawl.org", and processed by AllenAI.alpaca-spanish
BERTIN Alpaca Spanish
This dataset is a translation to Spanish of alpaca_data_cleaned.json, a clean version of the Alpaca dataset made at Stanford.
An earlier version used Facebook's NLLB 1.3B model, but the current version uses OpenAI's gpt-3.5-turbo, hence this dataset cannot be used to create models that compete in any way against OpenAI.
zenobia-instruct-hf
Zenobia Instruct
This dataset has been extracted from alvp/zenobia and alvp/stanzas, and parsed into a huggingface-friendly format so you can use apply_chat_template as explained on the Chat Templating documentation.
Example
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.1")
chat = [
{"role": "user", "content": "Escribe un terceto sobre la naturaleza en un paisaje nevado."},
{"role":… See the full description on the dataset page: https://huggingface.co/datasets/bertin-project/zenobia-instruct-hf.BOE-XSUM
BOE-XSUM Balanced Dataset - Reviewed and Cleaned
Description
The BOE 2025 Dataset is a collection of BOE articles with extreme summaries of them. This dataset has been carefully balanced and cleaned to ensure its quality and usefulness in natural language processing (NLP) tasks, primarily for evaluating generative models.
Read more in https://arxiv.org/abs/2509.24908
Dataset Content
The dataset is composed of the following subsets (splits):
train: Training… See the full description on the dataset page: https://huggingface.co/datasets/bertin-project/BOE-XSUM.bonanza-hf
Bonanza: Dataset de instrucciones en Español y Catalán
Este dataset combina múltiples fuentes para proporcionar instrucciones en español y catalán. Los datasets combinados son los siguientes:
OpenAssistant/oasst2
CohereForAI/aya_dataset
projecte-aina/RAG_Multilingual
bertin-project/alpaca-spanish
dariolopez/Llama-2-databricks-dolly-oasst1-es
projecte-aina/MentorESprojecte-aina/MentorCA
Descripción
Este conjunto de datos proporciona una rica colección de instrucciones… See the full description on the dataset page: https://huggingface.co/datasets/bertin-project/bonanza-hf.oasst2_es_instruct_hfThis is the Spanish subset from the OpenAssistant/oasst2 dataset.
The dataset has been extracted from the 2023-11-05_oasst2_ready.trees.jsonl.gz file to parse all the conversation trees and put it in a huggingface-friendly format so you can use apply_chat_template as explained on the Chat Templating documentation.
Example
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.1")
chat = [
{"role": "user"… See the full description on the dataset page: https://huggingface.co/datasets/bertin-project/oasst2_es_instruct_hf.practica-unirpractica-2BERTing