{
  "dataset_name": "real_human_conversations",
  "description": "Real human-to-human conversations from Reddit, forums, and public sources. Natural, uncensored dialogue across multiple languages.",
  "total_conversations": 2285855,
  "train_size": 2171562,
  "validation_size": 68575,
  "test_size": 45718,
  "total_size_mb": 1598.9878845214844,
  "sources": {
    "reddit_posts": 800000,
    "reddit_finnish": 300000,
    "reddit_danish": 299937,
    "reddit_confessions": 246133,
    "opensubtitles_japanese": 232232,
    "opensubtitles_welsh": 150000,
    "reddit_jokes": 96815,
    "reddit_aita": 90000,
    "reddit_questions": 53864,
    "korean_dialogue": 16874
  },
  "languages": {
    "en": 1286812,
    "fi": 300000,
    "da": 299937,
    "ja": 232232,
    "cy/en": 150000,
    "ko": 16874
  },
  "format": "parquet",
  "schema": {
    "conversation_id": "unique ID",
    "source": "data source",
    "language": "language code",
    "turns": "dialogue turns",
    "text": "plain text with speaker labels",
    "chatml": "ChatML format",
    "messages_json": "JSON messages array",
    "subreddit": "source community",
    "score": "community score",
    "created_utc": "timestamp"
  },
  "created_at": "2026-05-08 17:10:24.576132"
}