https://github.com/taoyds/typesql
TypeSQL: Knowledge-based Type-Aware Neural Text-to-SQL Generation
https://github.com/taoyds/typesql
Last synced: 5 months ago
JSON representation
TypeSQL: Knowledge-based Type-Aware Neural Text-to-SQL Generation
- Host: GitHub
- URL: https://github.com/taoyds/typesql
- Owner: taoyds
- Created: 2018-04-13T02:08:43.000Z (about 7 years ago)
- Default Branch: master
- Last Pushed: 2022-04-15T17:19:31.000Z (about 3 years ago)
- Last Synced: 2024-08-18T11:13:40.264Z (9 months ago)
- Language: Python
- Homepage:
- Size: 41 KB
- Stars: 112
- Watchers: 11
- Forks: 31
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- Awesome-Text2SQL - [code
README
## TypeSQL
Source code accompanying our NAACL 2018 paper:[TypeSQL: Knowledge-based Type-Aware Neural Text-to-SQL Generation
](https://arxiv.org/abs/1804.09769)
:+1: `03/20/2022`: **We open-sourced a simple but SOTA model (just T5) for the task! Please check out our code in the [UnifiedSKG repo](https://github.com/hkunlp/unifiedskg)!!**
#### Environment Setup
1. The code uses Python 2.7 and [Pytorch 0.2.0](https://pytorch.org/previous-versions/) GPU.
2. Install Python dependency: `pip install -r requirements.txt`
3. Install Pytorch 0.2.0: `conda install pytorch=0.2.0 cuda91 -c pytorch`. Replace cuda91 to whichever cuda version you have.
#### Download Data and Embeddings
1. Download the zip data file at the [Google Drive](https://drive.google.com/file/d/1CGIRCjwf2bgmWl3UyjY1yJpP4nU---Q0/view?usp=sharing), and put it in the root dir.
2. Download the pretrained [Glove](https://nlp.stanford.edu/data/wordvecs/glove.42B.300d.zip) and the [paraphrase embedding](https://drive.google.com/file/d/1iWTowxEG1-KZyq-fHP6cb6dNqMh4eHiN/view?usp=sharing) `para-nmt-50m/data/paragram_sl999_czeng.txt`. Put the unziped glove and para-nmt-50m folders in the root dir.
#### Train Models
1. To use knowledge graph types:
```
mkdir saved_model_kg
python train.py --sd saved_model_kg
```
2. To use DB content types:
```
mkdir saved_model_con
python train.py --sd saved_model_con --db_content 1
```
#### Test Models
1. Test Model with knowledge graph types:
```
python test.py --sd saved_model_kg
```
2. Test Model with knowledge graph types:
```
python test.py --sd saved_model_con --db_content 1
```
#### Get Data Types
1. Get a Google Knowledge Graph Search API Key by following the [link](https://developers.google.com/knowledge-graph/)
2. Search knowledge graph to get entities:
```
python get_kg_entities.py [Google freebase API Key] [input json file] [output json file]
```
3. Use detected knowledge graph entites and DB content to group questions and create type attributes in data files:
```
python data_process_test.py --tok [output json file generated at step 2] --table TABLE_FILE --out OUTPUT_FILE [--data_dir DATA_DIRECTORY] [--out_dir OUTPUT_DIRECTORY]
python data_process_train_dev.py --tok [output json file generated at step 2] --table TABLE_FILE --out OUTPUT_FILE [--data_dir DATA_DIRECTORY] [--out_dir OUTPUT_DIRECTORY]
```
#### Acknowledgement
The implementation is based on [SQLNet](https://github.com/xiaojunxu/SQLNet). Please cite it too if you use this code.