I’m seeing my runs being sent to LangSmith and successfully populated into my tracing project. However, the feedback/evaluator I’ve enabled — called missing_tool_call — has stopped running. Up until around 12pm 9/8, the feedback was working fine. After that, no logs are showing up and I can’t see any related errors.
Error message: {‘error’: ‘Evaluator failed to batch invoke {“detail”:“Service unavailable”}’}’
Happy to attach some runs as well.