M

Mrebel Large

Developed by Babelscape
REDFM is the multilingual version of REBEL, designed for multilingual relation extraction tasks, supporting relation triple extraction in 18 languages.
Downloads 573
Release Time : 6/12/2023

Model Overview

This model reformulates relation extraction as a sequence-to-sequence task, capable of extracting subject-relation-object triples from text and supporting multiple languages.

Model Features

Multilingual Support
Supports relation extraction in 18 languages, including major languages such as English, Chinese, and Spanish.
Sequence-to-Sequence Framework
Reframes relation extraction as a sequence-to-sequence problem, simplifying traditional pipeline approaches.
Typed Triples
Not only extracts entity relations but also identifies the types of entities and relations.

Model Capabilities

Multilingual text analysis
Relation triple extraction
Entity type recognition
Cross-lingual relation extraction

Use Cases

Knowledge Graph Construction
Automated Multilingual Knowledge Graph Construction
Extracts entity relations from multilingual texts to build or expand knowledge graphs.
Automatically identifies entities and their relations in text, forming structured knowledge.
Information Extraction
News Event Analysis
Extracts relationships between people, organizations, and locations from news texts.
Quickly understands key figures and relationship networks in news events.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase