Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Improving Test score using DIN SQL #40

Open
goyalpramod opened this issue Jun 17, 2023 · 1 comment
Open

Improving Test score using DIN SQL #40

goyalpramod opened this issue Jun 17, 2023 · 1 comment

Comments

@goyalpramod
Copy link
Contributor

Spider Is topped by DIN-SQL two aspects have not been explored in the mentioned model namely. DB content used and DSP. Implementation of the mentioned features in DIN-SQL may present substantial improvement in the test score.

Colab Notebook. This is the current Notebook in which the work is being pursued.

@goyalpramod
Copy link
Contributor Author

goyalpramod commented Jun 19, 2023

Current progress:

  • A breakdown of the DIN SQL model has been prepared and added to the notebook
  • Literature review of the DIN SQL paper has been made & a PR will be made soon to add the information

Problems encountered:

  • The model can not be shifted to gpt 3.5 to reduce the cost and time as tokens become the bottleneck
  • Cannot figure out how "db content used", can be used in this model.

Scope of Improvement:

  • The DIN SQL model does not have prompts for extra hard questions, they need to be framed and added specifically to the model
  • Ways to add DSP need to be explored

Papers:
Link The following paper talks about QA on knowledge database, This can be useful to understand how to implement DSP on DIN SQL (Nothing of substance was found)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant