| --- |
| language: |
| - en |
| tags: |
| - text2sql |
| datasets: |
| - splash |
| widget: |
| - text: "Give the name, population, and head of state for the country that has the largest area. || select name, population, continent from country order by surfacearea desc limit 1 || world_1 | country : name, population, headofstate, surfacearea || swap continent with head of state because it is not required." |
| --- |
| ## parkervg/destt5-text2sql |
|
|
| Fine-tuned weights for the text2sql model described in [Correcting Semantic Parses with Natural Language through Dynamic |
| Schema Encoding](https://arxiv.org/pdf/2305.19974.pdf), based on [t5-base](https://huggingface.co/t5-base). |
|
|
|
|
| ### Training Data |
|
|
| The model has been fine-tuned on the 7,481 training examples in the [SPLASH interactive semantic parsing dataset](https://github.com/MSR-LIT/Splash). |
|
|
| Rather than seeing the full database schema, it only received the filtered schema as predicted by the [destt5-schema-prediction model](https://huggingface.co/parkervg/destt5-schema-prediction) |
|
|
|
|
| ### Training Objective |
|
|
| This model was initialized with [t5-base](https://huggingface.co/t5-base) and fine-tuned with the text-to-text generation objective. |
|
|
| As this model works in the interactive setting, we utilize the standard text2sql features such as `question` and `db_schema`, in addition to `feedback` and `incorrect_parse`. |
|
|
| Importantly, the `[table]`, `[column]`, `[content]` features are expected to be the 'gold' schema items, as predicted by an initial auxiliary schema prediction model. |
|
|
| ``` |
| [question] || [incorrect_parse] || [db_id] | [table] : [column] ( [content] , [content] ) , [column] ( ... ) , [...] | [table] : ... | ... || [feedback] |
| ``` |
|
|
| The model then attempts to parse the corrected SQL query, using the filtered database schema items. This is prefaced by the `db_id`. |
|
|
| ``` |
| [db_id] | [sql] |
| ``` |
|
|
|
|
| ### Performance |
|
|
| When this model receives the serialized database schema as predicted by [destt5-schema-prediction](https://huggingface.co/parkervg/destt5-schema-prediction), it achieves 53.43% correction accuracy (exact-match) on the SPLASH test set. |
|
|
|
|
| ### References |
|
|
| 1. [Correcting Semantic Parses with Natural Language through Dynamic |
| Schema Encoding](https://arxiv.org/pdf/2305.19974.pdf) |
|
|
| 2. [DestT5 codebase](https://github.com/parkervg/destt5) |
|
|
| 3. [Speak to your Parser: Interactive Text-to-SQL with Natural Language Feedback](https://arxiv.org/pdf/2005.02539v2.pdf) |
|
|
|
|
| ### Citation |
|
|
| ```bibtex |
| @inproceedings{glenn2023correcting, |
| author = {Parker Glenn, Parag Pravin Dakle, Preethi Raghavan}, |
| title = "Correcting Semantic Parses with Natural Language through Dynamic Schema Encoding", |
| booktitle = "Proceedings of the 5th Workshop on NLP for Conversational AI", |
| publisher = "Association for Computational Linguistics", |
| year = "2023" |
| } |
| ``` |
|
|
