dataset_summary.json

{
  "source": "LiteFold/BFD",
  "source_sha": "61cd373acbdc6265b698ea70bd88c8dd8f0b4d27",
  "viewer_table_scope": "compressed archive byte-range index",
  "source_file": "bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt.tar.gz",
  "source_size_bytes": 291649557441,
  "chunk_size_gib": 1,
  "chunk_rows": 272,
  "splits": {
    "train": 255,
    "test": 17
  },
  "split_strategy": "deterministic sha256(index_id) % 10; bucket 0 is test, buckets 1-9 are train",
  "columns": [
    "index_id",
    "repo_id",
    "source_file",
    "source_sha",
    "source_format",
    "chunk_index",
    "byte_start",
    "byte_end_exclusive",
    "chunk_size_bytes",
    "total_size_bytes",
    "chunk_size_gib",
    "is_first_chunk",
    "is_last_chunk",
    "access_note",
    "split_bucket"
  ]
}