Hi, I’m currently trying to get the output column in eval/datasets to output something so that I can run the grader through it.
However, I can’t seem to get the assistant to actually output something after a tool call was invoked. So when I run the grader, it will error out because output is not present.
Things I’ve tried (same issue for all):
- Defining custom tools
- Adding function calls
- Connecting to custom mcp server (current)
Wondering if this is a bug or if anyone have any ideas on how can I resolve this?