picard VS lux

This is an active field of research. You might want to look at the main challenge dataset for it : https://yale-lily.github.io/spider. It would be interesting to use ChatGPT's model as a pre-processor and then feed its output in to a more finetuned model like PICARD.

We tried using OpenAI/Davinci for SQL query authoring, but it quickly became obvious that we are still really far from something the business could find value in. The state of the art as described below is nowhere near where we would need it to be:

https://yale-lily.github.io/spider

https://arxiv.org/abs/2109.05093

https://github.com/ElementAI/picard

To be clear, we haven't tried this on actual source code (i.e. procedural concerns), so I feel like this is a slightly different battle.

The biggest challenge I see is that the queries we would need the most assistance with are the same ones that are the rarest to come by in terms of training data. They are also incredibly specific in the edge cases, many time requiring subjective evaluation criteria to produce an acceptable outcome (i.e. recursive query vs 5k lines of unrolled garbage).

Hi again! This is a surprise, we chose Apache 2 precisely so that businesses could use our code without legal problems or costs, see https://github.com/ElementAI/picard/blob/main/LICENSE. The only things that are excluded are liability and warranty. It is what it is :)