Are these phenomena expected?

[wenxian-yang]

Thanks for the great work!
When testing the "ToolPlanner_Stage2_1107" model with the test data "data/test_sample/G3_query_100_opendomain.json", it was found that the tag and tagtrace results for all 100 cases were empty. Additionally, it was observed that in the code implementation, tagtrace is not included in the query passed to the root node of the DST for subsequent inference (see toolbench/inference/Downstream_tasks/rapidapi.py L1201 and L1246). Are these phenomena expected?
Thank you!