Evaluation: Fix evaluate API failure when trace.destination is set to none (Azure#38023)

* Handle the case where the customer manually run "pf config set trace.destination=none"

* fix the black issue

* update changelog

Author: ninghu

sdk/evaluation/azure-ai-evaluation/CHANGELOG.md

@@ -10,6 +10,7 @@
 
 ### Bugs Fixed
 - Non adversarial simulator works with `gpt-4o` models using the `json_schema` response format
+- Fix evaluate API failure when `trace.destination` is set to `none`
 
 ### Other Changes
 - Improved error messages for the `evaluate` API by enhancing the validation of input parameters. This update provides more detailed and actionable error descriptions.

sdk/evaluation/azure-ai-evaluation/azure/ai/evaluation/_evaluate/_evaluate.py

@@ -690,6 +690,11 @@ def _evaluate(  # pylint: disable=too-many-locals,too-many-statements
         )
 
     trace_destination: Optional[str] = pf_client._config.get_trace_destination()  # pylint: disable=protected-access
+
+    # Handle the case where the customer manually run "pf config set trace.destination=none"
+    if trace_destination and trace_destination.lower() == "none":
+        trace_destination = None
+
     target_run: Optional[Run] = None
 
     # Create default configuration for evaluators that directly maps