fix: evaluation docstring examples to use value=0 instead of value=None (#1520)

docs: fix Evaluation docstring examples to use value=0 instead of value=None

Update docstring examples in EvaluatorFunction and RunEvaluatorFunction
protocols to use value=0 instead of value=None, matching the type definition
which requires Union[int, float, str, bool] and does not allow None.

This change aligns with commit d11155e
which established that Evaluation score values cannot be None.

Changes:
- Update accuracy_evaluator example in EvaluatorFunction docstring
- Update llm_judge_evaluator error handling example
- Update average_accuracy example in RunEvaluatorFunction docstring
- Update accuracy_evaluator example in DatasetClient.run_experiment docstring

Author: Tkgmuuuu

langfuse/_client/datasets.py

@@ -286,7 +286,7 @@ def answer_questions(*, item, **kwargs):
 
             def accuracy_evaluator(*, input, output, expected_output=None, **kwargs):
                 if not expected_output:
-                    return {"name": "accuracy", "value": None, "comment": "No expected output"}
+                    return {"name": "accuracy", "value": 0, "comment": "No expected output"}
 
                 is_correct = output.strip().lower() == expected_output.strip().lower()
                 return {

langfuse/experiment.py

@@ -719,7 +719,7 @@ def __call__(
             ```python
             def accuracy_evaluator(*, input, output, expected_output=None, **kwargs):
                 if expected_output is None:
-                    return {"name": "accuracy", "value": None, "comment": "No expected output"}
+                    return {"name": "accuracy", "value": 0, "comment": "No expected output"}
 
                 is_correct = output.strip().lower() == expected_output.strip().lower()
                 return {
@@ -773,7 +773,7 @@ async def llm_judge_evaluator(*, input, output, expected_output=None, **kwargs):
                 except ValueError:
                     return {
                         "name": "llm_judge_quality",
-                        "value": None,
+                        "value": 0,
                         "comment": "Could not parse LLM judge score"
                     }
             ```
@@ -867,7 +867,7 @@ def average_accuracy(*, item_results, **kwargs):
                             accuracy_values.append(evaluation.value)
 
                 if not accuracy_values:
-                    return {"name": "avg_accuracy", "value": None, "comment": "No accuracy evaluations found"}
+                    return {"name": "avg_accuracy", "value": 0, "comment": "No accuracy evaluations found"}
 
                 avg = sum(accuracy_values) / len(accuracy_values)
                 return {