french document embedding EU AI Act Compliance Profile

Model Description:

french-document-embedding is an embedding model for documents in the French-English language, with a context length of up to 8096 tokens. This model is a specialized text-embedding model trained specifically for the French-English language. It is built upon gte-multilingual and trained using the [SimilarityLoss], Multi-Negative Ranking Loss, Matryoshka2dLoss and GISTEmbedLoss using guide model. This model embeds and converts long texts or documents into vectors with 786 dimensions, making it useful for vector databases serving semantic search or RAG (Retrieval-Augmented Generation).

Full Model Architecture

Usage:

Using this model becomes easy when you have sentence-transformers installed:

Then you can use the model like this:

from sentence_transformers import SentenceTransformer
sentences = ["Paris est une capitale de la France", "Paris is a capital of France"]



model = SentenceTransformer('dangvantuan/french-document-embedding', trust_remote_code=True)
embeddings = model.encode(sentences)
print(embeddings)

Evaluation

Citation

Your obligations as a french document embedding deployer

Track your progress

Sign up to track obligations, receive deadline reminders, and generate audit packages.

Tags

Languages

Library