MatchBERT

A BioBERT-based cross-encoder fine-tuned for complex pairwise matching tasks.

  • Base model: dmis-lab/biobert-base-cased-v1.1
  • Training data: 87.8M pairs (clinical trial matching)
  • Architecture: Cross-encoder with two-stage retrieval
  • Author: Simón Dhungana, The George Washington University

Paper forthcoming. Model weights will be released upon publication.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support