StyleECU

StyleECU is a style embedding model for Spanish, obtained by fine-tuning mStyleDistance on SynthSTEL-ES, a purpose-built Spanish contrastive dataset of 51,400 triplets covering 71 stylistic dimensions.

Model Description

StyleECU specializes the mStyleDistance embedding space toward stylistic phenomena most relevant to Spanish, including dialectal variation (voseo/tuteo), expressive morphology, syntactic complexity, and digital style.

Training

• Base model: StyleDistance/mstyledistance
• Training objective: TripletLoss (contrastive learning)
• Dataset: cespinr/SynthSTEL-ES
• Training size: 51,400 triplets
• Epochs: 2

Usage

from sentence_transformers import SentenceTransformer

model = SentenceTransformer("cespinr/StyleECU")
embeddings = model.encode(["Tu texto aquí"])

Evaluation

Evaluated on PAN author profiling tasks (Spanish):

Task	Base (mStyleDistance)	StyleECU	Δ
PAN 2018 – Gender prediction	baseline	+3 pp	+3 pp
PAN 2021 – Hate speech spreaders	0.70	0.81	+11 pp

Authors

César Espín-Riofrio — Researcher, University of Guayaquil, Ecuador & SINAI, University of Jaén, Spain | Director, Research Project FCI-036-2023, University of Guayaquil, Ecuador

Arturo Montejo-Ráez — Researcher, SINAI, University of Jaén, Spain

Steven Ramírez-Gurumendi, Gabriel Delgado-Gómez University of Guayaquil, Ecuador — Research Project FCI-036-2023

Citation

If you use this model, please cite:

Paper under review. Citation will be updated upon publication.

@misc{espinriofrio2026stylecu,
  author = {Espín-Riofrio, César and Montejo-Ráez, Arturo and
            Ramírez-Gurumendi, Steven and Delgado-Gómez, Gabriel},
  title  = {StyleECU: A Spanish Style Embedding Model},
  year   = {2026},
  url    = {https://huggingface.co/cespinr/StyleECU}
}