Question Answering on SQuAD2.0 dev

Metric: F1 (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Hide extra data

#	Model↕	F1▼	Extra Data	Paper	Date↕	Code
1	XLNet (single model)	90.6	No	XLNet: Generalized Autoregressive Pretraining fo...	2019-06-19	Code
2	XLNet+DSC	89.51	No	Dice Loss for Data-imbalanced NLP Tasks	2019-11-07	Code
3	RoBERTa (no data aug)	89.4	Yes	RoBERTa: A Robustly Optimized BERT Pretraining A...	2019-07-26	Code
4	ALBERT xxlarge	88.1	No	ALBERT: A Lite BERT for Self-supervised Learning...	2019-09-26	Code
5	SG-Net	87.9	No	SG-Net: Syntax-Guided Machine Reading Comprehens...	2019-08-14	Code
6	SpanBERT	86.8	No	SpanBERT: Improving Pre-training by Representing...	2019-07-24	Code
7	ALBERT xlarge	85.9	No	ALBERT: A Lite BERT for Self-supervised Learning...	2019-09-26	Code
8	SemBERT large	83.6	No	Semantics-aware BERT for Language Understanding	2019-09-05	Code
9	ALBERT large	82.1	No	ALBERT: A Lite BERT for Self-supervised Learning...	2019-09-26	Code
10	ALBERT base	79.1	No	ALBERT: A Lite BERT for Self-supervised Learning...	2019-09-26	Code
11	RMR + ELMo (Model-III)	74.8	No	Read + Verify: Machine Reading Comprehension wit...	2018-08-17	-
12	U-Net	74	No	U-Net: Machine Reading Comprehension with Unansw...	2018-10-12	Code
13	TinyBERT-6 67M	73.4	No	TinyBERT: Distilling BERT for Natural Language U...	2019-09-23	Code