Made it possible to get answers from litqa evaluations #760
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
dosubot
bot
added
size:M
jamesbraza
approved these changes
Dec 11, 2024
| UNSURE = "unsure" | ||
|
|
||
| @property | ||
| def answer(self) -> str | None: |
There was a problem hiding this comment.
Ha so basically this PR moves LitQAEvaluation to be basically an Enum/tuple hybrid.
I can tell from this PR that you want to just keep moving, but a better pattern would be something like a dataclass or BaseModel that contains an Evaluation. So to be more specific:
- Dataclass
LitQAEntryfor an LitQA question that containsanswer: str,ideal: str,selection: Evaluation - Moves all methods (e.g.
from_question/from_answer) to the dataclassLitQAEntry
Does that make sense?
Of course it's more of a breaking change, but you can remove the str | None awkwardness and have simpler code
Also, feel free to just say let's save for another PR
There was a problem hiding this comment.
Gonna leave this here: #761
Will see how things shake-out before refactor though
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I'm not super happy with the pattern, but I modified
LitQAEvaluationso that it stuffs the extracted answer into theStrEnumvalue. This makes it so we can recover the actual answers at the end, to help with consensus sampling.