Spider 1.0 test image

Yale Semantic Parsing and Text-to-SQL Challenge

What is Spider?

Spider is a large-scale complex and cross-domain semantic parsing and text-to-SQL dataset annotated by 11 Yale students. The goal of the Spider challenge is to develop natural language interfaces to cross-domain databases. It consists of 10,181 questions and 5,693 unique complex SQL queries on 200 databases with multiple tables covering 138 different domains. In Spider 1.0, different complex SQL queries and databases appear in train and test sets. To do well on it, systems must generalize well to not only new SQL queries but also new database schemas.
Why we call it "Spider"? It is because our dataset is complex and cross-domain like a spider crawling across mutiple complex(with many foreign keys) nests(databases). Spider Paper (EMNLP'18) Spider Post
Related challenges: multi-turn SParC and conversational CoSQL text-to-SQL tasks. SParC Challenge (ACL'19) CoSQL Challenge (EMNLP'19)

News

  • 11/15/2020 We will use Test Suite Accuracy as our official evaluation metric for Spider, SParC, and CoSQL. Please find the evaluation code from here. Also, Notice that Test results after May 02, 2020 are reported on the new release (collected some annotation errors).
  • 08/03/2020 Corrected "column_name" and "column_name_original" mismatches in 2 dbs ("scholar" and "formula_1") in tables.json, and reparsed SQL queries (this only affects some models (e.g. RATSQL) which use our parsed SQL as the SQL input). Please download the Spider dataset from this page again.
  • 06/07/2020 We corrected some annotation errors and label mismatches (not errors) in Spider dev and test sets (~4% of dev examples updated, click here for more details). Please download the Spider dataset from this page again.
  • 01/16/2020 For value prediction (in order to compute the execution accuracy), your model should be able to 1) copy from the question inputs, 2) retrieve from the database content (database content is available), or 3) generate numbers (e.g. 3 in "LIMIT 3").
  • 9/24/2019 (Min et al., EMNLP 2019) translated Spider to Chinese! Check out the Chinese challenge page.
  • 5/17/2019 Our paper SParC: Cross-Domain Semantic Parsing in Context with Salesforce Research was accepted to ACL 2019! It introduces the context-dependent version of the Spider challenge: SParC!
  • 5/17/2019 Please report any annotation errors here, we really appreciate your help and will update the data release in this summer!
  • 1/14/2019 The submission tutorial is out!.
  • 12/17/2018 We updated 7 sqlite database files (issue 14). Please download the Spider dataset from this page again.
  • 10/25/2018 The evaluation script and results were updated (issue 5). Please download the lastest versions of the script and papers. Also, please follow instructions in issue 3 to generate the latest SQL parsing results (fixed a bug).

Why Spider?

test image

As the above spider chart shows, Spider 1.0 is distinct from most of the previous semantic parsing tasks because:
  • ATIS, Geo, Academic: Each of them contains only a single database with a limited number of SQL queries, and has exact same SQL queries in train and test splits.
  • WikiSQL: The numbers of SQL queries and tables are significantly large. But all SQL queries are simple, and each database is only a simple table without any foreign key.
Spider 1.0 spans the largest area in the chart, making it the first complex and cross-domain semantic parsing and text-to-SQL dataset! Read more on the blog post.

Getting Started

The data is split into training, development, and unreleased test sets. Download a copy of the dataset (distributed under the CC BY-SA 4.0 license):

Spider Dataset Details of baseline models and evaluation script can be found on the following GitHub site: Spider GitHub Page

Once you have built a model that works to your expectations on the dev set, you can submit it to get official scores on the dev and a hidden test set. To preserve the integrity of test results, we do not release the test set to the public. Instead, we request you to submit your model so that we can run it on the test set for you. Here's a tutorial walking you through official evaluation of your model:

Submission Tutorial

Data Examples

Some examples look like the following:

test image

Have Questions or Want to Contribute ?

Ask us questions at our Github issues page or contact Tao Yu, Rui Zhang, or Michihiro Yasunaga.

We expect the dataset to evolve. We would greatly appreciate it if you could donate us your non-private databases or SQL queries for the project.

Acknowledgement

We thank Graham Neubig, Tianze Shi, Catherine Finegan-Dollak, and the anonymous reviewers for their precious comments on this project. Also, we thank Pranav Rajpurkar for giving us the permission to build this website based on SQuAD.

Our team at the summit of the East Rock park in New Haven (The pose is "NLseq2SQL"):

test image

Leaderboard - Execution with Values

Our current models do not predict any value in SQL conditions so that we do not provide execution accuracies. However, we encourage you to provide it in the future submissions. For value prediction, your model should be able to 1) copy from the question inputs, 2) retrieve from the database content (database content is available), or 3) generate numbers (e.g. 3 in "LIMIT 3"). Notice: Test results after May 02, 2020 are reported on the new release (collected some annotation errors).

Rank Model Test

1

Nov 24, 2020 		BRIDGE v2 + BERT(ensemble) (DB content used)

Salesforce Research

(Lin et al., EMNLP-Findings '20) code 		68.3

2

Nov 24, 2020 		BRIDGE v2 + BERT (DB content used)

Salesforce Research

(Lin et al., EMNLP-Findings '20) code 		64.3

3

May 30, 2020 		AuxNet + BART (DB content used)

Anonymous

 62.6

4

May 30, 2020 		BRIDGE + BERT (DB content used)

Salesforce Research

(Lin et al., EMNLP-Findings '20) code 		59.9

5

May 20, 2020 		GAZP + BERT (DB content used)

University of Washington & Facebook AI Research

(Zhong et al., EMNLP '20) 		53.5

Leaderboard - Exact Set Match without Values

For exact matching evaluation, instead of simply conducting string comparison between the predicted and gold SQL queries, we decompose each SQL into several clauses, and conduct set comparison in each SQL clause. Please refer to the paper and the Github page for more details. Notice: Test results after May 02, 2020 are reported on the new release (collected some annotation errors).

Rank Model Dev Test

1

Nov 19, 2020 		RAT-SQL + GraPPa + Adv (DB content used)

Anonymous

 75.5 70.5

2

Dec 25, 2020 		RATSQL + GraPPa + GP (DB content used)

OCFT Gamma Big Data Lab

(Zhao et al.,'21) 		72.8 69.8

2

Sep 08, 2020 		RATSQL + GAP (DB content used)

University of Waterloo & AWS AI Labs

(Shi et al., AAAI'21) code 		71.8 69.7

2

Aug 18, 2020 		RATSQL + GraPPa (DB content used)

Yale & Salesforce Research

(Yu et al., ICLR'21) code 		73.4 69.6

5

Nov 24, 2020 		BRIDGE v2 + BERT(ensemble) (DB content used)

Salesforce Research

(Lin et al., EMNLP-Findings '20) code 		71.1 67.5

6

May 02, 2020 		RATSQL v3 + BERT (DB content used)

Microsoft Research

(Wang and Shin et al., ACL '20) code 		69.7 65.6

7

Dec. 07, 2020 		DuoRAT + BERT (DB content used)

Anonymous

 - 65.4

7

Sep. 8, 2020 		YCSQL + BERT (DB content used)

Anonymous

 - 65.3

9

Nov 24, 2020 		BRIDGE v2 + BERT (DB content used)

Salesforce Research

(Lin et al., EMNLP-Findings '20) code 		70.0 65.0

10

Sep. 8, 2020 		ShadowGNN (DB content used)

Anonymous

 - 64.8

11

Sep. 8, 2020 		GP-RATSQL + BERT (DB content used)

Anonymous

 - 64.5

11

Nov. 25, 2020 		RATSQL-HPFT + BERT (DB content used)

Anonymous

 - 64.4

12

May 31, 2020 		AuxNet + BART (DB content used)

Anonymous

 70.0 61.9

13

Dec 13, 2019 		RATSQL v2 + BERT (DB content used)

Microsoft Research

(Wang and Shin et al., ACL '20) code 		65.8 61.9

14

May 31, 2020 		AuxNet + BART

Anonymous

 68.0 61.3

15

Feb 18, 2020 		RYANSQL v2 + BERT

Kakao Enterprise

(Choi et al., '20) 		70.6 60.6

15

Oct 19, 2020 		SmBoP + BART

Tel-Aviv University & Allen Institute for AI

(Rubin and Berant '20) 		66.0 60.5

17

Dec 18, 2019 		IRNet++ + XLNet (DB content used)

Anonymous

 65.5 60.1

18

May 30, 2020 		BRIDGE + BERT (DB content used)

Salesforce Research

(Lin et al., EMNLP-Findings '20) code 		65.5 59.2

19

Nov 12, 2019 		RYANSQL + BERT

Kakao Enterprise

(Choi et al., '20) 		66.6 58.2

20

Dec 13, 2019 		RATSQL v2 (DB content used)

Microsoft Research

(Wang and Shin et al., ACL '20) code 		62.7 57.2

21

Dec 13, 2019 		SLSQL + BERT + Data Annotation

National University of Singapore

(Lei and Wang et al., EMNLP '20) code 		60.8 55.7

22

Dec 13, 2019 		EditSQL+LSL + BERT

Anonymous

 57.9 55.2

23

June 24, 2019 		IRNet v2 + BERT

Microsoft Research Asia

 63.9 55.0

24

Sep 20, 2019 		GIRN + BERT

Anonymous

 60.2 54.8

24

May 19, 2019 		IRNet + BERT

Microsoft Research Asia

(Guo and Zhan et al., ACL '19) code 		61.9 54.7

26

Nov 4, 2019 		GNN + Bertrand-DR

Got It R&D

(Kelkar et al., '20) code 		57.9 54.6

27

Apr 8, 2020 		CNSQL

Anonymous

 58.0 54.0

28

Sep 19, 2019 		RATSQL

Anonymous

 60.6 53.7

29

Sep 1, 2019 		EditSQL + BERT

Yale University & Salesforce Research

(Zhang et al., EMNLP '19) code 		57.6 53.4

29

May 21, 2020 		GAZP + BERT

University of Washington & Facebook AI Research

(Zhong et al., EMNLP '20) 		- 53.3

31

May 21, 2020 		NatSQL v3

Anonymous

 - 53.2

32

May 28, 2020 		IRNET+ GNN

Anonymous

 - 49.6

33

June 24, 2019 		IRNet v2

Microsoft Research Asia

 55.4 48.5

34

Aug 30, 2019 		Global-GNN (DB content used)

Tel-Aviv University & Allen Institute for AI

(Bogin et al., EMNLP '19) code 		52.7 47.4

35

Dec 13, 2019 		LSL

Anonymous

 56.8 47.0

35

Apr 5, 2020 		GraphSQL

Anonymous

 52.8 46.9

37

May 19, 2019 		IRNet

Microsoft Research Asia

(Guo and Zhan et al., ACL '19) code 		53.2 46.7

37

Mar 17, 2020 		SG-IRNet

Anonymous

 - 46.6

39

Dec 13, 2019 		NatSQL v2

Anonymous

 52.0 46.4

40

June 11, 2019 		HSRNet

Anonymous

 51.5 45.6

41

June 12, 2019 		CFGN

Anonymous

 48.7 44.1

42

Aug 31, 2019 		NatSQL

Anonymous

 52.9 42.5

43

May 16, 2019 		GNN

Tel-Aviv University & Allen Institute for AI

(Bogin et al., ACL '19) code 		40.7 39.4

44

Feb 25, 2019 		SASeq

Anonymous

 40.8 37.4

45

May 30, 2019 		GrammarSQL

Allen Institute for AI

(Lin et al., '19) 		34.8 33.8

46

Sep 1, 2019 		EditSQL

Yale University & Salesforce Research

(Zhang et al., EMNLP '19) code 		36.4 32.9

47

Dec 13, 2019 		GuideSQL

Anonymous

 36.8 31.5

48

Sep 20, 2018 		SyntaxSQLNet + augment

Yale University

(Yu et al., EMNLP '18) code 		24.8 27.2

49

April 18, 2019 		RCSQL

SAP Labs Korea

(Lee, EMNLP'19) 		28.5 24.3

50

Sep 20, 2018 		SyntaxSQLNet

Yale University

(Yu et al., EMNLP '18) code 		18.9 19.7

51

Sep 20, 2018 		SQLNet

Shanghai Jiao Tong University (modified by Yale)

(Xu et al., '18) code 		10.9 12.4

52

Sep 20, 2018 		TypeSQL

Yale University

(Yu et al., NAACL '18) code 		8.0 8.2

53

Sep 20, 2018 		Seq2Seq + attention

University of Edinburgh (modified by Yale)

(Dong and Lapata, ACL '16) code 		1.8 4.8

Other papers used Spider (evaluated on the dev but not test set):
  1. (Min et al., EMNLP 2019), Westlake University, Spider in Chinese
  2. (Yao et al., EMNLP 2019), OSU & Facebook AI Research
  3. (Shaw et al., ACL 2019), Google
  4. (Shin et al., NeurlPS 2019), UC Berkeley & MSR
  5. (Weir et al., SIGMOD 2019), Brown University & TU Darmstadt
  6. (Baik et al., ICDE 2019), U of Michigan & IBM