dataset_infos.json

{
  "default": {
    "description": "RAG documents for MNLP",
    "features": {
      "title":   {"_type":"Value","dtype":"string"},
      "text":    {"_type":"Value","dtype":"string"},
      "source":  {"_type":"Value","dtype":"string"}
    },
    "splits": {
      "train": {
        "name": "train",
        "num_bytes":  0,
        "num_examples": 621689
      }
    }
  }
}