{
  "version": 1,
  "dataset": "ava-actions",
  "bucket_id": "TornadoLabs/staging",
  "inventory_observed_at": "2026-09-29T16:25:33.435526+00:00",
  "cap_bytes": 1000000000000,
  "reserve_bytes": 1000000000,
  "seed": "tornadolabs-public-v1",
  "method": "sha256-weighted-round-robin-by-source-count",
  "stratum_field": "split",
  "before_bytes": 90416053103,
  "after_bucket_bytes": 90416053103,
  "remove_bytes": 0,
  "before_videos": 151,
  "after_videos": 151,
  "retained_unrecognized_objects": 0,
  "strata": {
    "train": {
      "before_videos": 81,
      "after_videos": 81,
      "before_bytes": 50341317751,
      "after_bytes": 50341317751
    },
    "unknown": {
      "before_videos": 53,
      "after_videos": 53,
      "before_bytes": 29707773660,
      "after_bytes": 29707773660
    },
    "validation": {
      "before_videos": 17,
      "after_videos": 17,
      "before_bytes": 10366961692,
      "after_bytes": 10366961692
    }
  },
  "plan_sha256": "404a8b95f431bf684a568194f6a32a16ae00780608a8ff7740d7878660ece82c",
  "size_unit": "decimal bytes; 1 TB = 1,000,000,000,000 bytes",
  "sampling_limitations": "Selected from previously hosted source videos, not the full upstream corpus. Weighted round robin follows source counts within the available annotation strata; the byte constraint can favor smaller groups near the cap. No new language or duration balancing."
}