README.md

metadata

license: apache-2.0
pipeline_tag: text-generation
tags:
  - code
datasets:
  - semiotic/SynQL-KaggleDBQA-Train
language:
  - en
base_model:
  - google-t5/t5-3b

Model Card for T5-3B/SynQL-KaggleDBQA-Train-Run-02

• Developed by: Semiotic Labs
• Model type: [Text to SQL]
• License: [Apache-2.0]
• Finetuned from model: google-t5/t5-3b
• Dataset used for finetuning: semiotic/SynQL-KaggleDBQA-Train

Model Context

Example metadata can be found below, context represents the prompt that is presented to the model. Database schemas follow the encoding method proposed by Shaw et al (2020).

"query": "SELECT count(*) FROM singer",
"question": "How many singers do we have?",
"context": "How many singers do we have? | concert_singer | stadium : stadium_id, location, name, capacity, highest, lowest, average | singer : singer_id, name, country, song_name, song_release_year, age, is_male | concert : concert_id, concert_name, theme, stadium_id, year | singer_in_concert : concert_id, singer_id",
"db_id": "concert_singer",

Model Results

Evaluation set: KaggleDBQA/test

Evaluation metrics: [Execution Accuracy]

Model	Data	Run	Execution Accuracy
T5-3B	semiotic/SynQL-KaggleDBQA	00	0.3514
T5-3B	semiotic/SynQL-KaggleDBQA	01	0.3514
T5-3B	semiotic/SynQL-KaggleDBQA	02	0.3514