SParC: Yale & Salesforce Semantic Parsing and Text-to-SQL in Context Challenge

What is SParC?

Nov 12, 2024: We have released Spider 2.0 full paper, data and code. Follow the guideline to submit your scores to the leaderboard!

Aug 28, 2024: The early access version of Spider 2.0 (a more realistic and challenging text-to-SQL task) is now available! We expect to release the whole dataset in 1-2 weeks. As this is a preliminary release, there may be errors. Your feedback would be invaluable in refining the dataset!

SParC is a dataset for cross-domain Semantic Parsing in Context. It is the context-dependent/multi-turn version of the Spider task, a complex and cross-domain text-to-SQL challenge. SParC consists of 4,298 coherent question sequences (12k+ unique individual questions annotated with SQL queries annotated by 14 Yale students), obtained from user interactions with 200 complex databases over 138 domains.

XLANG Lab for Building LLM/VLM Agents SParC Paper (ACL'19) SParC Post

Related Works from XLANG Lab: Spider 2.0 Text-to-SQL ('24) Spider2-V ('24) OSWorld ('24) DS-1000 Challenge (ICML'23) Binder Framework (ICLR '23) UnifiedSKG Framework (EMNLP'22) Spider Chanllenge (EMNLP'18) CoSQL Chanllenge (EMNLP'19)

News

11/12/2024 We have released Spider 2.0 full paper, data and code. Follow the guideline to submit your scores to the leaderboard.
08/28/2024 The early access version of Spider 2.0 (a more realistic and challenging text-to-SQL task) is now available! As this is a preliminary release, there may be errors. Your feedback would be invaluable in refining the dataset!
07/15/2024 Spider 2.0-vision (Benchmarking Multimodal Agents on Automating Data Science and Engineering Workflows) is out! Spider 2.0-SQL (much more realistic and challenging than Spider 1.0!) will be released in August.
08/10/2023 Please check out XLANG Lab for Building LLM/VLM Agents!
11/20/2022 Please check out our recent work DS-1000: A Natural and Reliable Benchmark for Data Science Code Generation. Please check out examples, data, and code on the DS-1000 project site!!
10/18/2022 Please check out our recent work Binder: an easy but sota neural-symbolic built on GPT-3 Codex & SQL/Python interpreter. It injects GPT-3 Codex prompt API calls in programming languages! Please check out Binder demo, code, paper, and video on the Binder project site!!
02/15/2022 Please check out our recent work UnifiedSKG: Unifying and Multi-Tasking Structured Knowledge Grounding with Text-to-Text Language Models. We open-sourced simple but SOTA/strong models for 21 tasks including text-to-SQL! Please check out our code in the UnifiedSKG repo!!
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.
5/17/2019 Our paper SParC: Cross-Domain Semantic Parsing in Context with Salesforce Research was accepted to ACL 2019!

Why SParC?

SParC is built upon the Spider dataset. Comparing to other existing context-dependent semantic parsing/text-to-SQL datasets such as ATIS, it demonstrates:

complex contextual dependencies (annotated by 15 Yale computer science students)
has greater semantic diversity due to complex coverage of SQL logic patterns in the Spider dataset.
requires generalization to new domains due to its cross-domain nature and the unseen databasest time.

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):

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

Another example: test image

Have Questions or Want to Contribute ?

Ask us questions at our Github issues page or contact Tao Yu, Rui Zhang, or Xi Victoria Lin.

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 Tianze Shi and the anonymous reviewers for their precious comments on this project and Melvin Gruesbeck for designing the nice example illustrations. Also, we thank Pranav Rajpurkar for giving us the permission to build this website based on SQuAD. .

Part of our SParC team at YINS:

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").

Rank	Model	Question Match	Interaction Match
1 Jun 4, 2022	RASAT + PICARD SJTU LUMIA & Netmind.AI (Qi et al., EMNLP'22) code	74.0	52.6
2 May 24, 2020	TreeSQL V2 + BERT Anonymous	48.5	21.6
3 May 21, 2020	GAZP + BERT University of Washington & Facebook AI Research (Zhong et al., EMNLP '20)	44.6	19.7

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.

Rank	Model	Question Match	Interaction Match
1 Feb 14, 2022	STAR Alibaba DAMO & SIAT (Cai and Li et al., EMNLP-Findings '22) code demo	67.4	46.6
2 Jun 4, 2022	RASAT + PICARD SJTU LUMIA & Netmind.AI (Qi et al., EMNLP'22) code	67.7	45.2
3 Apr 27, 2022	CQR-SQL Tencent Cloud Xiaowei (Xiao et al.,'22)	68.2	44.4
4 Oct 8, 2021	RAT-SQL-TC + GAP Meituan & PKU (Li et al.,'21)	65.7	43.2
5 Oct 18, 2021	HIE-SQL + GraPPa Alibaba DAMO (Zheng et al. ACL-Findings '22)	64.6	42.9
6 Sep. 21, 2020	RAT-SQL + SCoRe Yale & Microsoft Research & PSU (Yu et al. ICLR '21)	62.4	38.1
7 Oct 21, 2020	WaveSQL+BERT Anonymous	58.7	33.3
8 July 08, 2020	R²SQL + BERT Alibaba DAMO (Hui et al. AAAI '21) code	55.8	30.8
9 May 26, 2020	IGSQL + BERT Peking University (Cai et al. EMNLP '20) code	51.2	29.5
10 Jun. 02, 2020	MIE + BERT Anonymous	49.6	27.1
11 May 04, 2020	SubTreeSQL + BERT Anonymous	47.4	25.5
12 Sep 1, 2019	EditSQL + BERT Yale University & Salesforce Research (Zhang et al. EMNLP '19) code	47.9	25.3
13 May 03, 2020	TreeSQL V2 + BERT Anonymous	48.1	25.0
14 May 22, 2020	MH-LTA + BERT Anonymous	48.5	24.7
15 Jan 15, 2020	TreeSQL + BERT Anonymous	46.3	24.3
16 May 21, 2020	GAZP + BERT University of Washington & Facebook AI Research (Zhong et al., EMNLP '20)	45.9	23.5
17 Feb 13, 2020	ConcatSQL + BERT Anonymous	46.3	22.4
18 Apr 21, 2021	MemCE UoE (Jain et al., TACL '21)	40.3	16.7
19 Feb 13, 2020	ConcatSQL Anonymous	39.0	16.3
20 Dec 13, 2019	GuideSQL Anonymous	34.4	13.1
21 May 17, 2019	CD-Seq2Seq Yale University & Salesforce Research (Yu et al. ACL '19) code	23.2	7.5
22 May 17, 2019	SyntaxSQL-con Yale University (Yu et al. EMNLP '18) code	20.2	5.2

SParC 1.0

Yale & Salesforce Semantic Parsing and Text-to-SQL in Context Challenge