{
  "split": "train-00012",
  "output": "data/tokenized-v2-20b/initial/train-00012.bin",
  "input_files": 1,
  "documents_seen": 50000,
  "documents_written": 50000,
  "tokens": 46697792,
  "bytes": 93395584,
  "dtype": "uint16",
  "input": "data/clean-v2-full/train-00012.jsonl"
}