task: text-classification

dataset:
  name: intent-detection-benchmark
  split: test

prediction:
  type: label

metrics:
  - name: accuracy
  - name: f1

primary_metric: f1

description: Intent detection benchmark for search queries.