{
  "dataset": "stard",
  "language": "zh",
  "num_documents": 55348,
  "num_queries": 1543,
  "num_train_queries": 1235,
  "num_test_queries": 308,
  "num_relevance_judgments": 2717,
  "num_train_judgments": 2205,
  "num_test_judgments": 512,
  "avg_relevant_docs_per_query": 1.7608554763447828,
  "note": "dev split used as test; queries not in train/dev assigned to train"
}