Added LLM-based intent validation in ModelBuilder#158
Added LLM-based intent validation in ModelBuilder#158AumPatel1 wants to merge 2 commits intoplexe-ai:mainfrom
Conversation
_validate_intent() method that performs basic checks (non-empty, length) and LLM-based validation
Greptile OverviewGreptile SummaryAdded intent validation to Key changes:
The implementation follows good error handling patterns with explicit Confidence Score: 4/5
Important Files ChangedFile Analysis
|
There was a problem hiding this comment.
Additional Comments (1)
-
plexe/model_builder.py, line 92 (link)syntax: comment says "Optional feedback" but
feedback: stris required - should befeedback: Optional[str] = Noneor remove "Optional" from comment
1 file reviewed, 1 comment
|
Closing this PR as stale. Unfortunately it was opened at a time when we were not actively maintaining this project, and now relates to code that no longer exists. However, I do see value in the notion of validating whether the input intent sufficiently specifies the problem. If you'd like to reopen and update this PR for the current version of the project, feel free to. Thanks for being interesting in contributing! |
In ModelBuilder added a _validate_intent() method that performs basic checks and LLM-based validation ,to check that intent passed is not too long, empty and valid