TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/BERT-FP

BERT-FP

Reported on 21 benchmarks across 2 tasks · 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 Processing20 results

  • Conversational Response SelectiononDouban
    MAP
    0.644
    best: 0.651 (SEMSOL(W/o utterances))
  • Conversational Response SelectiononDouban
    MRR
    0.68
    best: 0.688 (Uni-Enc+BERT-FP)
  • Conversational Response SelectiononDouban
    P@1
    0.512
    best: 0.518 (Uni-Enc+BERT-FP)
  • Conversational Response SelectiononDouban
    R10@1
    0.324
    best: 0.33 (SEMSOL)
  • Conversational Response SelectiononDouban
    R10@2
    0.542
    best: 0.557 (Uni-Enc+BERT-FP)
  • Conversational Response SelectiononDouban
    R10@5
    0.87
    best: 0.877 (SEMSOL(W/o utterances))
  • Conversational Response SelectiononRRS
    MAP
    0.702
  • Conversational Response SelectiononRRS
    MRR
    0.712
    best: 0.715 (SA-BERT+BERT-FP)
  • Conversational Response SelectiononRRS
    P@1
    0.543
    best: 0.555 (SA-BERT+BERT-FP)
  • Conversational Response SelectiononRRS
    R10@1
    0.488
    best: 0.497 (SA-BERT+BERT-FP)
  • Conversational Response SelectiononRRS
    R10@2
    0.708
  • Conversational Response SelectiononRRS
    R10@5
    0.927
    best: 0.931 (SA-BERT+BERT-FP)
  • Conversational Response SelectiononRRS Ranking Test
    NDCG@3
    0.609
    best: 0.679 (Poly-encoder)
  • Conversational Response SelectiononRRS Ranking Test
    NDCG@5
    0.709
    best: 0.765 (Poly-encoder)
  • Conversational Response SelectiononUbuntu Dialogue (v1, Ranking)
    R10@1
    0.911
    best: 0.918 (Dial-MAE)
  • Conversational Response SelectiononUbuntu Dialogue (v1, Ranking)
    R10@2
    0.962
    best: 0.965 (BERT-FP+EDHNS)
  • Conversational Response SelectiononUbuntu Dialogue (v1, Ranking)
    R10@5
    0.994
  • Conversational Response SelectiononE-commerce
    R10@1
    0.87
    best: 0.957 (BERT-FP+EDHNS)
  • Conversational Response SelectiononE-commerce
    R10@2
    0.956
    best: 0.986 (BERT-FP+EDHNS)
  • Conversational Response SelectiononE-commerce
    R10@5
    0.993
    best: 0.997 (BERT-FP+EDHNS)

Speech1 result

  • DialogueonHarry Potter Dialogue Dataset
    Recall 10@1· 2022-11-13
    0.259
    SOTA
    Large Language Models Meet Harry Potter: A Bilingual Dataset for Aligning Dialogue Agents with CharactersarXiv:2211.06869