Open Zeen0 opened 2 months ago
Would you write your environment? Thank you!
v1.8.4
class Labels(abc.ABC):
label_class = ExportedLabel
column = "labels"
fields: Tuple[str, ...] = ("example", "label") # To boost performance
def __init__(self, examples: QuerySet[ExportedExample], user=None):
self.label_groups = defaultdict(list)
labels = self.label_class.objects.filter(example__in=examples)
if user:
labels = labels.filter(user=user)
for label in labels.select_related(*self.fields):
self.label_groups[label.example.id].append(label)
def find_by(self, example_id: int) -> Dict[str, List[ExportedLabel]]:
return {self.column: self.label_groups[example_id]}
It looks like this might exist in the form of user binary, but can't seem to find any settings in /admin/ or the UI to activate it.
Feature description
When exporting annotations made by multiple annotators on a single project, there is no way to discern who made which annotation. This would allow for projects to be more centralized for downstream tasks.
I notice that in the admin panel the names user IDs are associated with spans, but not in data export.