Two-Step+BERT-base

Reported on 2 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing2 results

Relation ExtractiononDocRED
F1· 2019-09-26
53.92
best: 67.53 (DREEAM)
SOTA
Fine-tune Bert for DocRED with Two-step Process arXiv:1909.11898
Relation ExtractiononDocRED
Ign F1· 2019-09-26
54.42
best: 65.47 (DREEAM)
Fine-tune Bert for DocRED with Two-step Process arXiv:1909.11898