List your company

Near-duplicate identification on C4 Sampled 50M documents

2,766,436Number of DuplicatesSEDD
2,516,507.362,581,392.682,646,2782,711,163.32Jan 2, 2025

Evaluation results

MethodLinks
SEDD2025.012,766,4362,486,83598.4
GPU baseline (NeMo Curator)2025.012,526,120——