diff --git a/spacy/pipeline/span_finder.py b/spacy/pipeline/span_finder.py
index af4109c76..f11bcf793 100644
--- a/spacy/pipeline/span_finder.py
+++ b/spacy/pipeline/span_finder.py
@@ -111,6 +111,7 @@ def span_finder_score(examples: Iterable[Example], **kwargs) -> Dict[str, Any]:
     )
     kwargs.setdefault("has_annotation", lambda doc: key in doc.spans)
     kwargs.setdefault("allow_overlap", True)
+    kwargs.setdefault("labeled", False)
     scores = Scorer.score_spans(examples, **kwargs)
     scores.pop(f"{kwargs['attr']}_per_type", None)
     return scores
@@ -215,6 +216,7 @@ class SpanFinder(TrainablePipe):
                         break
                     elif self.min_length <= span_length:
                         doc.spans[self.spans_key].append(doc[start : end + 1])
+            offset += len(doc)
 
     def update(
         self,
diff --git a/spacy/tests/pipeline/test_span_finder.py b/spacy/tests/pipeline/test_span_finder.py
index 4caa3a33f..9c6b6c3e0 100644
--- a/spacy/tests/pipeline/test_span_finder.py
+++ b/spacy/tests/pipeline/test_span_finder.py
@@ -234,7 +234,7 @@ def test_overfitting_IO():
     scores = nlp.evaluate(train_examples)
     assert f"span_finder_{span_finder.spans_key}_f" in scores
     # It's not perfect 1.0 F1 because it's designed to overgenerate for now.
-    assert scores[f"span_finder_{span_finder.spans_key}_f"] == 0.4
+    assert scores[f"span_finder_{span_finder.spans_key}_p"] == 0.75
     assert scores[f"span_finder_{span_finder.spans_key}_r"] == 1.0
 
     # also test that the spancat works for just a single entity in a sentence