This dataset mirrors the Facility Support Analyzer dataset used in the Llama Prompt Ops tutorial. Each example is a facility-related message with gold labels for urgency, sentiment, and a set of service categories.

Tasks

Urgency (single label): low | medium | high

Sentiment (single label): positive | neutral | negative

Categories (multi-label): one or more from:

emergency_repair_services

routine_maintenance_requests… See the full description on the dataset page: https://huggingface.co/datasets/nlile/facility-support-analyzer.