Datasets:

---

GT-111

/

intern_a1_v3_eef

Like 0

Tidy of GT-111/intern_a1_v3_eef per results/hub-alignment-plan.md.

franka's head camera declares shape: [null, 640, 3] — a shape no consumer can allocate — and its video_info block has no video.height; its MP4 measures 640×360, so the declaration becomes [3, 360, 640]. genie1 declares [360, 640, 3] heads and [480, 848, 3] wrists. Every child writes its video metadata under video_info, which the loader ignores, and every child's Parquet carries an observation.image_mask column (fixed_size_list<bool>[3], one flag per camera) that meta/info.json does not declare. This makes the declarations concrete and channel-first, moves them to info, declares the mask, replaces invalid_transition_semantics with invalid_transitions: "masked", adds fps_semantics: "physical", and gives each child the action_units its own measured gripper range supports (franka {"gripper": "fraction", "position": "m"}; genie1 {"position": "m"}; lift2 {"gripper": "m", "position": "m"}; split_aloha {"gripper": "m", "position": "m"}); where that range fits neither a metre opening nor an opening fraction, the unit is left unstated rather than invented. No value and no video changes.

Renamed keys

• collection_manifest.json and dataset_receipt.json: schema: "xq_collection_manifest_v2" / "xq_dataset_receipt_v2" -> schema: "collection_manifest" / "dataset_receipt".

• meta/action_contract.json's schema_id: "eef_absolute_next_observation_wxyz_v2" is dropped, and so is the same value in collection_manifest.json's and dataset_receipt.json's top-level action_contract field; what it encoded is carried by action_units, by each child's meta/action_contract.json (whose path the manifest lists as action_contract_path) and by the card sentence.

• codebase_version: "v3.0" is unchanged: that field names LeRobot's own format version, which the naming rule exempts.