chore: Add prompt eval to task type #629
Conversation
dcbfac4 to
73a0098
Compare
|
|
||
| @staticmethod | ||
| def get_valid_tasks() -> List[str]: | ||
| return list(map(lambda x: x.value, TaskType)) | ||
| def get_valid_tasks() -> List["TaskType"]: |
There was a problem hiding this comment.
let's keep this returning a string
There was a problem hiding this comment.
I thought about it but it doesn't have any downstream impact, which is why I felt comfortable changing it. Curious why you think we should match the signature?
There was a problem hiding this comment.
ahh cool, i didn't know the enum in an f string would work, but since it does, i'm cool with it
| def get_valid_tasks() -> List[str]: | ||
| return list(map(lambda x: x.value, TaskType)) | ||
| def get_valid_tasks() -> List["TaskType"]: | ||
| """Tasks that are valid for dataquality.""" |
There was a problem hiding this comment.
if we do it this way, let's add more to this docstring to explain the context of what "valid" means and why prompt evaluation isn't "valid"
| return [ | ||
| task_type | ||
| for task_type in TaskType | ||
| if task_type not in [TaskType.prompt_evaluation] |
There was a problem hiding this comment.
i would maybe suggest adding a new static method something like get_logger_task_types or get_dq_task_types or something along those lines that excludes prompt eval instead of hijacking this fn
73a0098 to
388361d
Compare
388361d to
053824e
Compare
Just adding the new task type to keep
dataqualityaligned.