Question Answering On Squad11

评估指标

EM
F1

评测结果

各个模型在此基准测试上的表现结果

Paper TitleRepository
{ANNA} (single model)90.62295.719--
LUKE 483M-95.4LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention
LUKE (single model)90.20295.379LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention
LUKE (single model)90.20295.379--
XLNet (single model)89.89895.080XLNet: Generalized Autoregressive Pretraining for Language Understanding
XLNet (single model)89.89895.080--
XLNET-123 (single model)89.64694.930--
XLNET-123++ (single model)89.85694.903--
XLNET-123+ (single model)89.70994.859--
SpanBERT (single model)88.83994.635--
SpanBERT (single model)88.894.6SpanBERT: Improving Pre-training by Representing and Predicting Spans
BERTSP (single model)88.91294.584--
Unnamed submission by NMC88.91294.584--
BERT+WWM+MT (single model)88.65094.393--
Tuned BERT-1seq Large Cased (single model)87.46593.294--
BERT-LARGE (Ensemble+TriviaQA)87.493.2BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
BERT (ensemble)87.43393.160BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
BART (TextBox 2.0)-93.04TextBox 2.0: A Text Generation Library with Pre-trained Language Models
LinkBERT (large)87.4592.7LinkBERT: Pretraining Language Models with Document Links
BERT+MT (single model)86.45892.645--
0 of 213 row(s) selected.
Question Answering On Squad11 | SOTA | HyperAI超神经