This is the official implementation of the paper "RESDSQL: Decoupling Schema Linking and Skeleton Parsing for Text-to-SQL" (AAAI 2023).
If this repository could help you, please cite the following paper:
@inproceedings{li2022resdsql,
author = {Haoyang Li and Jing Zhang and Cuiping Li and Hong Chen},
title = "RESDSQL: Decoupling Schema Linking and Skeleton Parsing for Text-to-SQL",
booktitle = "AAAI",
year = "2023"
}
Update (2023.3.13): We evaluated our method on a diagnostic evaluation benchmark, Dr.Spider, which contains 17 test sets to measure the robustness of Text-to-SQL parsers under different perturbation perspectives.
We introduce a new Text-to-SQL parser, RESDSQL (Ranking-enhanced Encoding plus a Skeleton-aware Decoding framework for Text-to-SQL), which attempts to decoulpe the schema linking and the skeleton parsing to reduce the difficuty of Text-to-SQL. More details can be found in our paper. All experiments are conducted on a single NVIDIA A100 80G GPU.
We evaluate RESDSQL on five benchmarks: Spider, Spider-DK, Spider-Syn, Spider-Realistic, and Dr.Spider. We adopt two metrics: Exact-set-Match accuracy (EM) and EXecution accuracy (EX). Let's look at the following numbers:
On Spider:
Model | Dev EM | Dev EX | Test EM | Test EX | Google Drive/OneDrive | Baidu Netdisk |
---|---|---|---|---|---|---|
RESDSQL-3B+NatSQL | 80.5% | 84.1% | 72.0% | 79.9% | OneDrive link | Link (pwd: 4r98) |
RESDSQL-3B | 78.0% | 81.8% | - | - | Google Drive link | Link (pwd: sc62) |
RESDSQL-Large+NatSQL | 76.7% | 81.9% | - | - | Google Drive link | Link (pwd: 7iyq) |
RESDSQL-Large | 75.8% | 80.1% | - | - | Google Drive link | Link (pwd: q58k) |
RESDSQL-Base+NatSQL | 74.1% | 80.2% | - | - | Google Drive link | Link (pwd: pyxf) |
RESDSQL-Base | 71.7% | 77.9% | - | - | Google Drive link | Link (pwd: wuek) |
On Spider-DK, Spider-Syn, and Spider-Realistic:
Model | DK EM | DK EX | Syn EM | Syn EX | Realistic EM | Realistic EX |
---|---|---|---|---|---|---|
RESDSQL-3B+NatSQL | 53.3% | 66.0% | 69.1% | 76.9% | 77.4% | 81.9% |
On Dr.Spider's perturbation sets: Following Dr.Spider, we only report EX for each post-perturbation set and choose PICARD and CodeX as our baseline methods.
Perturbation set | PICARD | CodeX | RESDSQL-3B | RESDSQL-3B+NatSQL |
---|---|---|---|---|
DB-Schema-synonym | 56.5% | 62.0% | 63.4% | 68.2% |
DB-Schema-abbreviation | 64.7% | 68.6% | 64.5% | 70.0% |
DB-DBcontent-equivalence | 43.7% | 51.6% | 40.3% | 40.1% |
NLQ-Keyword-synonym | 66.3% | 55.5% | 67.5% | 72.4% |
NLQ-Keyword-carrier | 82.7% | 85.2% | 86.7% | 83.5% |
NLQ-Column-synonym | 57.2% | 54.7% | 57.4% | 63.1% |
NLQ-Column-carrier | 64.9% | 51.1% | 69.9% | 63.9% |
NLQ-Column-attribute | 56.3% | 46.2% | 58.8% | 71.4% |
NLQ-Column-value | 69.4% | 71.4% | 73.4% | 76.6% |
NLQ-Value-synonym | 53.0% | 59.9% | 53.8% | 53.2% |
NLQ-Multitype | 57.1% | 53.7% | 60.1% | 60.7% |
NLQ-Others | 78.3% | 69.7% | 77.3% | 79.0% |
SQL-Comparison | 68.0% | 66.9% | 70.2% | 82.0% |
SQL-Sort-order | 74.5% | 57.8% | 79.7% | 85.4% |
SQL-NonDB-number | 77.1% | 89.3% | 83.2% | 85.5% |
SQL-DB-text | 65.1% | 72.4% | 67.8% | 74.3% |
SQL-DB-number | 85.1% | 79.3% | 85.4% | 88.8% |
Average | 65.9% | 64.4% | 68.2% | 71.7% |
Notice: We also employed the modified test suite script (see this issue) to evaluate the model-generated results, but obtained the same numbers as above. Nevertheless, we suggest that further work should use their modified script to evaluate Dr.Spider.
Create a virtual anaconda environment:
conda create -n your_env_name python=3.8.5
Active it and install the cuda version Pytorch:
conda install pytorch==1.11.0 torchvision==0.12.0 torchaudio==0.11.0 cudatoolkit=11.3 -c pytorch
Install other required modules and tools:
pip install -r requirements.txt
pip install https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-2.2.0/en_core_web_sm-2.2.0.tar.gz
python nltk_downloader.py
Create several folders:
mkdir eval_results
mkdir models
mkdir tensorboard_log
mkdir third_party
mkdir predictions
Clone evaluation scripts:
cd third_party
git clone https://github.com/ElementAI/spider.git
git clone https://github.com/ElementAI/test-suite-sql-eval.git
mv ./test-suite-sql-eval ./test_suite
cd ..
Download Spider, Spider-DK, Spider-Syn, Spider-Realistic, and Dr.Spider from data (add Dr.Spider in 2023.3.13) and database and then unzip them:
unzip data.zip
unzip database.zip
Notice: Dr.Spider has been preprocessed following the instructions on its Github page.
The evaluation results can be easily reproduced through our released scripts and checkpionts.
First, you should download T5 checkpoints from the table shown above. Then, you must download our cross-encoder checkpoints because our method is two-stage. For SQL and NatSQL, we train two cross-encoder because the classification labels of them are slightly different. Here are links:
Cross-encoder Checkpoints | Google Drive | Baidu Netdisk |
---|---|---|
Cross-encoder (for SQL) | Link | Link (pwd: dr62) |
Cross-encoder (for NatSQL) | Link | Link (pwd: 18w8) |
After downloading and unpacking all checkpoints, the models
folder should be:
├── models
│ ├── text2natsql_schema_item_classifier
│ ├── text2natsql-t5-3b
│ │ ├── checkpoint-61642
│ │ └── checkpoint-78302
│ ├── text2natsql-t5-base
│ │ └── checkpoint-14352
│ ├── text2natsql-t5-large
│ │ └── checkpoint-21216
│ ├── text2sql_schema_item_classifier
│ ├── text2sql-t5-3b
│ │ └── checkpoint-103292
│ ├── text2sql-t5-base
│ │ └── checkpoint-39312
│ └── text2sql-t5-large
│ └── checkpoint-30576
The inference scripts are located in scripts/inference
. Concretely, infer_text2natsql.sh
represents RESDSQL-{Base, Large, 3B}+NatSQL, infer_text2sql.sh
represents RESDSQL-{Base, Large, 3B}. For example, you can run the inference of RESDSQL-3B+NatSQL on Spider's dev set via:
sh scripts/inference/infer_text2natsql.sh 3b spider
The first argument (model scale) can be selected from [base, large, 3b]
and the second argument (dataset name) can be selected from [spider, spider-realistic, spider-syn, spider-dk, DB_schema_synonym, DB_schema_abbreviation, DB_DBcontent_equivalence, NLQ_keyword_synonym, NLQ_keyword_carrier, NLQ_column_synonym, NLQ_column_carrier, NLQ_column_attribute, NLQ_column_value, NLQ_value_synonym, NLQ_multitype, NLQ_others, SQL_comparison, SQL_sort_order, SQL_NonDB_number, SQL_DB_text, SQL_DB_number]
.
The predicted SQL queries are recorded in predictions/{dataset_name}/{model_name}/pred.sql
.
We also provide scripts in scripts/text2natsql
and scripts/text2sql
to train our framework on Spider's training set and evaluate on Spider's dev set.
RESDSQL-{Base, Large, 3B}+NatSQL
# Step1: preprocess dataset
sh scripts/train/text2natsql/preprocess.sh
# Step2: train cross-encoder
sh scripts/train/text2natsql/train_text2natsql_schema_item_classifier.sh
# Step3: prepare text-to-natsql training and development set for T5
sh scripts/train/text2natsql/generate_text2natsql_dataset.sh
# Step4: fine-tune T5-3B (RESDSQL-3B+NatSQL)
sh scripts/train/text2natsql/train_text2natsql_t5_3b.sh
# Step4: (or) fine-tune T5-Large (RESDSQL-Large+NatSQL)
sh scripts/train/text2natsql/train_text2natsql_t5_large.sh
# Step4: (or) fine-tune T5-Base (RESDSQL-Base+NatSQL)
sh scripts/train/text2natsql/train_text2natsql_t5_base.sh
RESDSQL-{Base, Large, 3B}
# Step1: preprocess dataset
sh scripts/train/text2sql/preprocess.sh
# Step2: train cross-encoder
sh scripts/train/text2sql/train_text2sql_schema_item_classifier.sh
# Step3: prepare text-to-sql training and development set for T5
sh scripts/train/text2sql/generate_text2sql_dataset.sh
# Step4: fine-tune T5-3B (RESDSQL-3B)
sh scripts/train/text2sql/train_text2sql_t5_3b.sh
# Step4: (or) fine-tune T5-Large (RESDSQL-Large)
sh scripts/train/text2sql/train_text2sql_t5_large.sh
# Step4: (or) fine-tune T5-Base (RESDSQL-Base)
sh scripts/train/text2sql/train_text2sql_t5_base.sh
During training, the cross-encoder (the first stage) always keeps the best checkpoint, but T5 (the second stage) keeps all the intermediate checkpoints, because different test sets may achieve the best Text-to-SQL performance on different checkpoints. Therefore, given a test set, we need to evaluate all the intermediate checkpoints and compare their performance to find the best checkpoint. The evaluation results of checkpoints are saved in eval_results
.
Therefore, for Spider-DK, Spider-Syn, and Spider-Realistic, we can evaluate checkpoints of RESDSQL-3B+NatSQL with the scripts provided in scripts/evaluate_robustness
. Here is an example for Spider-DK:
# Step1: preprocess Spider-DK
sh scripts/evaluate_robustness/preprocess_spider_dk.sh
# Step2: Run evaluation on Spider-DK
sh scripts/evaluate_robustness/evaluate_on_spider_dk.sh
We would thanks to Hongjin Su and Tao Yu for their help in evaluating our method on Spider's test set. We would also thanks to PICARD (paper, code), NatSQL (paper, code), Spider (paper, dataset), Spider-DK (paper, dataset), Spider-Syn (paper, dataset), Spider-Realistic (paper, dataset), and Dr.Spider (paper, dataset) for their interesting work and open-sourced code and dataset.