fix(experiments): move evals out of root span (#1437)

Author: hassiebp

langfuse/_client/client.py

@@ -2864,50 +2864,50 @@ async def _process_experiment_item(
                 )
                 raise e
 
-            # Run evaluators
-            evaluations = []
+        # Run evaluators
+        evaluations = []
 
-            for evaluator in evaluators:
-                try:
-                    eval_metadata: Optional[Dict[str, Any]] = None
+        for evaluator in evaluators:
+            try:
+                eval_metadata: Optional[Dict[str, Any]] = None
 
-                    if isinstance(item, dict):
-                        eval_metadata = item.get("metadata")
-                    elif hasattr(item, "metadata"):
-                        eval_metadata = item.metadata
+                if isinstance(item, dict):
+                    eval_metadata = item.get("metadata")
+                elif hasattr(item, "metadata"):
+                    eval_metadata = item.metadata
 
-                    eval_results = await _run_evaluator(
-                        evaluator,
-                        input=input_data,
-                        output=output,
-                        expected_output=expected_output,
-                        metadata=eval_metadata,
+                eval_results = await _run_evaluator(
+                    evaluator,
+                    input=input_data,
+                    output=output,
+                    expected_output=expected_output,
+                    metadata=eval_metadata,
+                )
+                evaluations.extend(eval_results)
+
+                # Store evaluations as scores
+                for evaluation in eval_results:
+                    self.create_score(
+                        trace_id=trace_id,
+                        observation_id=span.id,
+                        name=evaluation.name,
+                        value=evaluation.value,  # type: ignore
+                        comment=evaluation.comment,
+                        metadata=evaluation.metadata,
+                        config_id=evaluation.config_id,
+                        data_type=evaluation.data_type,  # type: ignore
                     )
-                    evaluations.extend(eval_results)
-
-                    # Store evaluations as scores
-                    for evaluation in eval_results:
-                        self.create_score(
-                            trace_id=trace_id,
-                            observation_id=span.id,
-                            name=evaluation.name,
-                            value=evaluation.value,  # type: ignore
-                            comment=evaluation.comment,
-                            metadata=evaluation.metadata,
-                            config_id=evaluation.config_id,
-                            data_type=evaluation.data_type,  # type: ignore
-                        )
 
-                except Exception as e:
-                    langfuse_logger.error(f"Evaluator failed: {e}")
+            except Exception as e:
+                langfuse_logger.error(f"Evaluator failed: {e}")
 
-            return ExperimentItemResult(
-                item=item,
-                output=output,
-                evaluations=evaluations,
-                trace_id=trace_id,
-                dataset_run_id=dataset_run_id,
-            )
+        return ExperimentItemResult(
+            item=item,
+            output=output,
+            evaluations=evaluations,
+            trace_id=trace_id,
+            dataset_run_id=dataset_run_id,
+        )
 
     def _create_experiment_run_name(
         self, *, name: Optional[str] = None, run_name: Optional[str] = None