README.md

metadata

license: eupl-1.1
pipeline_tag: sentence-similarity

EUBERT Cross-Encoder

EUBERT MLM finetuned as cross-encoder for the 24 languages of 🇪🇺 via the sentence-transformers training set: https://sbert.net/datasets/AllNLI.tsv.gz

from sentence_transformers import CrossEncoder
model = CrossEncoder('EuropeanParliament/eubert_cross-encoder_v1', max_length=512)
scores = model.predict([('Query', 'Paragraph1'), ('Query', 'Paragraph2') , ('Query', 'Paragraph3')])

Author:

• Sébastien Campion sebastien.campion@europarl.europa.eu