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/mT5

mT5

Reported on 47 benchmarks across 8 tasks · 4 papers · 4 SOTA

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

Natural Language Processing28 results

  • Data-to-Text GenerationonXAlign
    BLEU4· 2022-02-01
    25
    best: 29.27 (Fact-aware embedding with mT5)
    SOTA
    XAlign: Cross-lingual Fact-to-Text Alignment and Generation for Low-Resource LanguagesarXiv:2202.00291
  • Cross-LingualonXTREME
    Sentence-pair Classification· 2020-10-22
    89.8
    best: 91 (Turing ULR v6)
    SOTA
    mT5: A massively multilingual pre-trained text-to-text transformerarXiv:2010.11934
  • Cross-Lingual TransferonXTREME
    Sentence-pair Classification· 2020-10-22
    89.8
    best: 91 (Turing ULR v6)
    SOTA
    mT5: A massively multilingual pre-trained text-to-text transformerarXiv:2010.11934
  • Question AnsweringonTweetQA
    BLEU-1· 2021-05-28
    70.8
    best: 72 (ByT5 (small))
    ByT5: Towards a token-free future with pre-trained byte-to-byte modelsarXiv:2105.13626
  • Question AnsweringonTweetQA
    ROUGE-L· 2021-05-28
    74.3
    best: 75.7 (ByT5)
    ByT5: Towards a token-free future with pre-trained byte-to-byte modelsarXiv:2105.13626
  • Extreme SummarizationonGEM-XSum
    BLEU score· 2021-05-28
    14.3
    best: 15.3 (ByT5)
    ByT5: Towards a token-free future with pre-trained byte-to-byte modelsarXiv:2105.13626
  • Data-to-Text GenerationonWebNLG ru
    METEOR· 2021-02-02
    0.18
    best: 0.613 (mBART)
    The GEM Benchmark: Natural Language Generation, its Evaluation and MetricsarXiv:2102.01672
  • Data-to-Text GenerationonWebNLG en
    METEOR· 2021-02-02
    0.287
    best: 0.462 (mBART)
    The GEM Benchmark: Natural Language Generation, its Evaluation and MetricsarXiv:2102.01672
  • Cross-LingualonXTREME
    Avg· 2020-10-22
    40.9
    best: 85.5 (Turing ULR v6)
    mT5: A massively multilingual pre-trained text-to-text transformerarXiv:2010.11934
  • Cross-LingualonXTREME
    Question Answering· 2020-10-22
    73.6
    best: 52.5 (Anonymous5)
    mT5: A massively multilingual pre-trained text-to-text transformerarXiv:2010.11934
  • Cross-Lingual TransferonXTREME
    Avg· 2020-10-22
    40.9
    best: 85.5 (Turing ULR v6)
    mT5: A massively multilingual pre-trained text-to-text transformerarXiv:2010.11934
  • Cross-Lingual TransferonXTREME
    Question Answering· 2020-10-22
    73.6
    best: 52.5 (Anonymous5)
    mT5: A massively multilingual pre-trained text-to-text transformerarXiv:2010.11934
  • Abstractive Text SummarizationonWITS
    BERTScore
    80.73
  • Abstractive Text SummarizationonWITS
    ROUGE-1
    40.6
    best: 42.32 (BART-IT)
  • Abstractive Text SummarizationonWITS
    ROUGE-2
    26.9
    best: 28.83 (BART-IT)
  • Abstractive Text SummarizationonWITS
    ROUGE-L
    37.43
    best: 38.84 (BART-IT)
  • Abstractive Text Summarizationonvietnews
    Rouge-1
    58.05
    best: 67.8 (Kết quả nghiên cứu)
  • Abstractive Text Summarizationonvietnews
    Rouge-2
    26.76
    best: 34.24 (ViT5 large)
  • Abstractive Text Summarizationonvietnews
    Rouge-L
    37.38
    best: 43.55 (ViT5 large)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Il Post
    BERTScore
    74.69
    best: 75.86 (mBART)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Il Post
    ROUGE-1
    35.04
    best: 38.91 (mBART)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Il Post
    ROUGE-2
    17.41
    best: 21.41 (mBART)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Il Post
    ROUGE-L
    28.68
    best: 32.08 (mBART)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Fanpage
    # Parameters
    390
    best: 610 (mBART)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Fanpage
    BERTScore
    72.77
    best: 73.4 (mBART)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Fanpage
    ROUGE-1
    34.13
    best: 36.52 (mBART)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Fanpage
    ROUGE-2
    15.76
    best: 17.52 (mBART)
  • Abstractive Text SummarizationonAbstractive Text Summarization from Fanpage
    ROUGE-L
    24.84
    best: 26.14 (mBART)

Knowledge Base16 results

  • Text SummarizationonWITS
    BERTScore
    80.73
  • Text SummarizationonWITS
    ROUGE-1
    40.6
    best: 42.32 (BART-IT)
  • Text SummarizationonWITS
    ROUGE-2
    26.9
    best: 28.83 (BART-IT)
  • Text SummarizationonWITS
    ROUGE-L
    37.43
    best: 38.84 (BART-IT)
  • Text Summarizationonvietnews
    Rouge-1
    58.05
    best: 67.8 (Kết quả nghiên cứu)
  • Text Summarizationonvietnews
    Rouge-2
    26.76
    best: 34.24 (ViT5 large)
  • Text Summarizationonvietnews
    Rouge-L
    37.38
    best: 43.55 (ViT5 large)
  • Text SummarizationonAbstractive Text Summarization from Il Post
    BERTScore
    74.69
    best: 75.86 (mBART)
  • Text SummarizationonAbstractive Text Summarization from Il Post
    ROUGE-1
    35.04
    best: 38.91 (mBART)
  • Text SummarizationonAbstractive Text Summarization from Il Post
    ROUGE-2
    17.41
    best: 21.41 (mBART)
  • Text SummarizationonAbstractive Text Summarization from Il Post
    ROUGE-L
    28.68
    best: 32.08 (mBART)
  • Text SummarizationonAbstractive Text Summarization from Fanpage
    # Parameters
    390
    best: 610 (mBART)
  • Text SummarizationonAbstractive Text Summarization from Fanpage
    BERTScore
    72.77
    best: 73.4 (mBART)
  • Text SummarizationonAbstractive Text Summarization from Fanpage
    ROUGE-1
    34.13
    best: 36.52 (mBART)
  • Text SummarizationonAbstractive Text Summarization from Fanpage
    ROUGE-2
    15.76
    best: 17.52 (mBART)
  • Text SummarizationonAbstractive Text Summarization from Fanpage
    ROUGE-L
    24.84
    best: 26.14 (mBART)

Adversarial3 results

  • Text GenerationonXAlign
    BLEU4· 2022-02-01
    25
    best: 29.27 (Fact-aware embedding with mT5)
    SOTA
    XAlign: Cross-lingual Fact-to-Text Alignment and Generation for Low-Resource LanguagesarXiv:2202.00291
  • Text GenerationonWebNLG ru
    METEOR· 2021-02-02
    0.18
    best: 0.613 (mBART)
    The GEM Benchmark: Natural Language Generation, its Evaluation and MetricsarXiv:2102.01672
  • Text GenerationonWebNLG en
    METEOR· 2021-02-02
    0.287
    best: 0.462 (mBART)
    The GEM Benchmark: Natural Language Generation, its Evaluation and MetricsarXiv:2102.01672