Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.
Sources
Badge
Host
Last check
Result
From source
huggingface.co
2026-11-15 03:48 UTC
passed
Model card
From the publisher at commit c23e4c6. Shown as plain text with links; images open at the publisher.
Current pre-trained cross-lingual sentence encoders approaches use sentence-level objectives only. This can lead to loss of information, especially for tokens, which then degrades the sentence representation. We propose MEXMA, a novel approach that integrates both sentence-level and token-level objectives. The sentence representation in one language is used to predict masked tokens in another language, with both the sentence representation and all tokens directly updating the encoder. We show that adding token-level objectives greatly improves the sentence representation quality across several tasks. Our approach outperforms current pre-trained cross-lingual sentence encoders on bi-text mining as well as several downstream tasks. We also analyse the information encoded in our tokens, and how the sentence representation is built from them.
You use this model as you would any other XLM-RoBERTa model, taking into account that the "pooler" has not been trained, so you should use the CLS the encoder outputs directly as your sentence representation:
@inproceedings{janeiro-etal-2025-mexma,
title = "{MEXMA}: Token-level objectives improve sentence representations",
author = "Janeiro, Jo{\~a}o Maria and
Piwowarski, Benjamin and
Gallinari, Patrick and
Barrault, Loic",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.acl-long.1168/",
doi = "10.18653/v1/2025.acl-long.1168",
pages = "23960--23995",
ISBN = "979-8-89176-251-0",
abstract = "Cross-lingual sentence encoders (CLSE) create fixed-size sentence representations with aligned translations. Current pre-trained CLSE approaches use sentence-level objectives only. This can lead to loss of information, especially for tokens, which then degrades the sentence representation. We propose MEXMA, a novel approach that integrates both sentence-level and token-level objectives. The sentence representation in one language is used to predict masked tokens in another language, with both the sentence representation and *all tokens directly update the encoder*. We show that adding token-level objectives greatly improves the sentence representation quality across several tasks. Our approach outperforms current pre-trained cross-lingual sentence encoders on bitext mining as well as several downstream tasks. We also analyse the information encoded in our tokens, and how the sentence representation is built from them."
}
Files
12 files · 2.26 GBWeights 2.24 GB (safetensors). Sizes and sha256 from the signed manifest.
import os
os.environ["HF_ENDPOINT"] = "https://freeweights.allternit.com"
from transformers import AutoModel, AutoTokenizer
tok = AutoTokenizer.from_pretrained("facebook/MEXMA", revision="c23e4c68ecc94a32c483fed8e947ccd0caaa3e58")
model = AutoModel.from_pretrained("facebook/MEXMA", revision="c23e4c68ecc94a32c483fed8e947ccd0caaa3e58", device_map="auto")
sentence-transformersEmbed text with sentence-transformers.
import os
os.environ["HF_ENDPOINT"] = "https://freeweights.allternit.com"
from sentence_transformers import SentenceTransformer
model = SentenceTransformer("facebook/MEXMA", revision="c23e4c68ecc94a32c483fed8e947ccd0caaa3e58")
vectors = model.encode(["Hello"])
unsloth 
mlx-community