Closed FruVirus closed 8 months ago
Hi @FruVirus,
Thanks for your kind words and for taking an interest in our project!
ValidateAndImprove
Operation. For each thought of the previous operations, it tries to improve the thought until it is valid (or the number of tries reaches the limit). You could implement a similar operation quite easily for continuous refinements until a certain score is reached. I hope this clears things up! Let me know if you have more questions.
Is the path generation still static (done before execution)? This sounds very limiting. Any plans to improve?
Here is a dynamic GoT implementation of your code: https://github.com/JayceNing/DGoT/blob/a0c364ad94bd4110c83485cd50f2e4ea7bbba9a5/graph_of_thoughts/operations/operations.py#L482
The repo also links to a paper.
Consider merging.
Hi there,
First of all, great work on implementing GoT and the publication of your paper---really appreciate you guys making your source code publicly available!
Operation
s effectively halt their execution for paths that fail to meet some criteria?Operation
? Or does the paper only mean that one can refine overThought
s and notOperation
s?