metadata
license: apache-2.0
language:
- fr
- en
base_model:
- BAAI/bge-m3
pipeline_tag: text-ranking
library_name: sentence-transformers
tags:
- evalllm2026
- sentence-transformers
- cross-encoder
- reranker
- loss:LambdaLoss
cea-list-ia/evalllm2026-reranker-geonames
This model was trained by the CEA-LIST to participate in the evalLLM2026 challenge.
Model Description
- Trained by: CEA-LIST
- Task: Entity Linking (GeoNames)
- Type: Reranker (Cross-Encoder)
- Training Data: Wikipedia
This reranker is a cross-encoder designed to improve the candidates selected by an embedding model.
Usage
Query Prefix
To use this model, you should prompt it with the following query prefix:
Represent this geographical sentence for retrieving relevant GeoNames terms:
More Information
For more details, please refer to the GitHub repository.