Restructure as multi-task DOMINO-Extended catalog

Migrate the initial adjust_bottle online-follow collection into the task/collector/version/shard layout and add catalog, collection metadata, and versioned provenance.

README.md

@@ -1,5 +1,5 @@
 ---
-pretty_name: DOMINO Online-Follow Adjust Bottle
+pretty_name: DOMINO Extended
 language:
 - en
 license: apache-2.0
@@ -16,85 +16,48 @@ tags:
 - video
 ---
 
-# DOMINO Online-Follow: Adjust Bottle
+# DOMINO Extended
 
-This dataset contains **50** successful `adjust_bottle` demonstrations for the
-ALOHA-AgileX embodiment in the DOMINO simulator.  Unlike the original
-precomputed-interception expert, the bottle first moves independently from outside
-the camera view and the expert repeatedly observes, approaches, follows, grasps,
-lifts, and completes the original upright placement task.  The controller does not
-receive a future object trajectory or a precomputed interception point.
+DOMINO Extended is a versioned collection of curated DOMINO-format robot
+demonstrations spanning multiple tasks, scenarios, collection methods, and motion
+profiles.  Raw HDF5 archives are canonical; derived LeRobot data and runtime caches
+are intentionally kept out of this repository.
 
-## Data summary
+Current catalog: **1 shards / 50 episodes**.
 
-- Episodes: **50** (25 left-arm, 25 right-arm)
-- Object assets: model 13 (**25**) and model 16 (**25**)
-- Recorded control frames: **16,651**
-- Cameras: `cam_high`, `cam_left_wrist`, `cam_right_wrist`
-- Camera frames fully decoded during audit: **66,604**
-- Nominal training FPS: **15**
-- Object speed: 0.040--0.117 m/s
-  (mean 0.0964, median 0.1010)
-- Entry angle: -7.4--7.4 degrees
-- Unique scene seeds: **30**
-- Unique scene-motion combinations: **50**
-- Mean demonstration duration: 10.61 seconds
+## Collections
 
-All 50 HDF5 episodes, instructions, trace files, and videos passed the repository
-quality checker.  All camera frames and videos were fully decoded; timestamps are
-strictly increasing; the inactive arm remains fixed; and no wrong-arm language
-labels were found.  Full per-episode provenance and the acceptance audit are in
-`summary.json` inside the archive.
+| Collection ID | Task | Scenario | Collector | Episodes | Speed (m/s) | Details |
+|---|---|---|---|---:|---:|---|
+| `adjust_bottle/online_follow/v1/clean/follow_fast/shard-000` | `adjust_bottle` | `clean` | `online_follow/v1` | 50 | 0.040--0.117 | [metadata](./metadata/adjust_bottle/online_follow/v1/clean_follow_fast_shard-000.json) |
 
-## Repository structure
+The machine-readable index is [`catalog.json`](./catalog.json).  Each entry records
+the artifact path, SHA-256 digest, collection parameters, and metadata path.
 
-```text
-dataset/
-  adjust_bottle/
-    aloha-agilex_clean_follow_fast_50.zip
-provenance/
-  collect_data_dynamic_follow.py
-  demo_dynamicvla_follow_dataset.yml
-  adjust_bottle.py
-  adjust_bottle_instructions.json
-manifest.json
-LICENSE
-```
-
-The ZIP follows the public DOMINO naming/layout convention and expands to:
+## Repository layout
 
 ```text
-aloha-agilex_clean_follow_fast/
-  data/          # episode0.hdf5 ... episode49.hdf5
-  instructions/  # 100 seen + 100 unseen language variants per episode
-  metrics/       # causal object/robot traces and quality measurements
-  video/         # one preview video per episode
-  scene_info.json
-  seed.txt
-  summary.json
+dataset/<task>/<collector>/<version>/<embodiment>_<scenario>_<motion>_shard-NNN_<episodes>.zip
+metadata/<task>/<collector>/<version>/<scenario>_<motion>_shard-NNN.json
+provenance/collections/<task>/<collector>/<version>/
+catalog.json
 ```
 
-Archive size: **0.56 GiB**  
-SHA-256: `bffbe20ba8be9e34b81a4e6c11d2435e6640b17918b85a23346f4568460104fe`
-
-## Download
+## Download one collection
 
 ```bash
-hf download HanningHanning/DOMINO-Online-Follow   dataset/adjust_bottle/aloha-agilex_clean_follow_fast_50.zip   --repo-type dataset --local-dir DOMINO-Online-Follow
-
-unzip DOMINO-Online-Follow/dataset/adjust_bottle/aloha-agilex_clean_follow_fast_50.zip -d data/adjust_bottle/
+hf download HanningHanning/DOMINO-Extended   dataset/adjust_bottle/online_follow/v1/aloha-agilex_clean_follow_fast_shard-000_50.zip   --repo-type dataset --local-dir DOMINO-Extended
 ```
 
+## Current online-follow collection
 
-
-## Training formats
-
-The archive is the canonical DOMINO HDF5 release.  It was also validated after
-conversion to LeRobot v2.1 (16,601 usable PUMA frames at
-15 FPS).  Derived LeRobot files and optical-flow/object-grounding
-caches are intentionally excluded to avoid duplicating images and publishing
-machine-specific runtime caches.  The conversion scripts are provided by PUMA in
-the DOMINO repository.
+The first collection contains 50 successful `adjust_bottle` demonstrations.  The
+bottle starts outside the camera view, moves independently, and the expert repeatedly
+observes, approaches, follows, grasps, lifts, and completes the original upright
+placement task.  It does not receive a future object trajectory or a precomputed
+interception point.  All HDF5 episodes, language instructions, traces, camera frames,
+and preview videos passed the repository quality audit; see the collection metadata
+for full statistics.
 
 ## License and attribution
 

catalog.json

@@ -0,0 +1,37 @@
+{
+  "schema_version": 1,
+  "name": "DOMINO Extended",
+  "description": "Curated multi-task, multi-scenario DOMINO-format manipulation data.",
+  "entries": [
+    {
+      "id": "adjust_bottle/online_follow/v1/clean/follow_fast/shard-000",
+      "task": "adjust_bottle",
+      "scenario": "clean",
+      "collector": "online_follow",
+      "collector_version": "v1",
+      "motion_profile": "follow_fast",
+      "embodiment": "aloha-agilex",
+      "shard": 0,
+      "episodes": 50,
+      "speed_mps": {
+        "min": 0.04,
+        "q10": 0.06881156437735318,
+        "q25": 0.09598322523604268,
+        "median": 0.1009865548537678,
+        "q75": 0.10734210861455262,
+        "q90": 0.11192400520920132,
+        "max": 0.11731706113377513,
+        "mean": 0.09644049489283639
+      },
+      "artifact": {
+        "path": "dataset/adjust_bottle/online_follow/v1/aloha-agilex_clean_follow_fast_shard-000_50.zip",
+        "size_bytes": 603007635,
+        "sha256": "bffbe20ba8be9e34b81a4e6c11d2435e6640b17918b85a23346f4568460104fe",
+        "file_count": 203,
+        "internal_root": "aloha-agilex_clean_follow_fast"
+      },
+      "metadata_path": "metadata/adjust_bottle/online_follow/v1/clean_follow_fast_shard-000.json",
+      "quality_status": "passed"
+    }
+  ]
+}

manifest.json → metadata/adjust_bottle/online_follow/v1/clean_follow_fast_shard-000.json

@@ -1,11 +1,14 @@
 {
   "schema_version": 1,
-  "dataset_name": "DOMINO Online-Follow Adjust Bottle",
+  "id": "adjust_bottle/online_follow/v1/clean/follow_fast/shard-000",
   "format": "DOMINO HDF5 archive",
   "task": "adjust_bottle",
+  "scenario": "clean",
+  "collector": "online_follow",
+  "collector_version": "v1",
+  "motion_profile": "follow_fast",
   "embodiment": "aloha-agilex",
-  "condition": "clean",
-  "variant": "follow_fast",
+  "shard": 0,
   "episodes": 50,
   "source_bytes": 1047858870,
   "episode_group_counts": {
@@ -15,32 +18,140 @@
     "video": 50
   },
   "artifact": {
-    "path": "dataset/adjust_bottle/aloha-agilex_clean_follow_fast_50.zip",
+    "path": "dataset/adjust_bottle/online_follow/v1/aloha-agilex_clean_follow_fast_shard-000_50.zip",
     "size_bytes": 603007635,
     "sha256": "bffbe20ba8be9e34b81a4e6c11d2435e6640b17918b85a23346f4568460104fe",
     "file_count": 203,
     "internal_root": "aloha-agilex_clean_follow_fast"
   },
   "provenance_files": {
-    "provenance/collect_data_dynamic_follow.py": {
+    "provenance/collections/adjust_bottle/online_follow/v1/collector.py": {
       "size_bytes": 66930,
       "sha256": "4a501aeac4ab2643d360fd1ef1815656e969ef9e38049466ec0ac510d52add40"
     },
-    "provenance/demo_dynamicvla_follow_dataset.yml": {
+    "provenance/collections/adjust_bottle/online_follow/v1/config.yml": {
       "size_bytes": 5139,
       "sha256": "70a5f92d88006b27bdc6578df5fd21f6c4dc5a341eb595f1f09b2effc5a7750d"
     },
-    "provenance/adjust_bottle.py": {
+    "provenance/collections/adjust_bottle/online_follow/v1/environment.py": {
       "size_bytes": 8544,
       "sha256": "bd122764af3f2f3e8b236280983e8949506c5e96b4b9040529812d61ee48f115"
     },
-    "provenance/adjust_bottle_instructions.json": {
+    "provenance/collections/adjust_bottle/online_follow/v1/instructions.json": {
       "size_bytes": 2890,
       "sha256": "3168108cde17a45c0671bda463a46559c4a3e1b0f914324a28c2999b505cf7eb"
     }
   },
   "source_summary_sha256": "7cb74255a3380c5346b548c544a2d89013ec98c5683b3b9d0eef12e7195ceede",
   "source_finalized_at_utc": "2026-07-28T15:28:44.229534+00:00",
+  "dataset_statistics": {
+    "arms": {
+      "left": 25,
+      "right": 25
+    },
+    "object_models": {
+      "13": 25,
+      "16": 25
+    },
+    "arm_model_combinations": {
+      "left/13": 12,
+      "right/13": 13,
+      "right/16": 12,
+      "left/16": 13
+    },
+    "speed_mps": {
+      "min": 0.04,
+      "q10": 0.06881156437735318,
+      "q25": 0.09598322523604268,
+      "median": 0.1009865548537678,
+      "q75": 0.10734210861455262,
+      "q90": 0.11192400520920132,
+      "max": 0.11731706113377513,
+      "mean": 0.09644049489283639
+    },
+    "speed_threshold_counts": {
+      ">=0.060": 46,
+      ">=0.080": 43,
+      ">=0.090": 41,
+      ">=0.100": 34,
+      ">=0.105": 19,
+      ">=0.110": 7,
+      ">=0.115": 4
+    },
+    "entry_angle_degrees": {
+      "min": -7.407407407407407,
+      "q10": -6.222222222222222,
+      "q25": -4.296296296296297,
+      "median": -0.8888888888888893,
+      "q75": 3.2592592592592595,
+      "q90": 5.688888888888888,
+      "max": 7.4074074074074066,
+      "mean": -0.34370370370370407
+    },
+    "offscreen_start_abs_x_m": {
+      "min": 0.56032,
+      "q10": 0.5630400000000001,
+      "q25": 0.56808,
+      "median": 0.5776,
+      "q75": 0.5856,
+      "q90": 0.59216,
+      "max": 0.5983999999999999,
+      "mean": 0.5778048
+    },
+    "action_start_abs_x_m": {
+      "min": 0.4406122448979592,
+      "q10": 0.44404081632653064,
+      "q25": 0.44566326530612244,
+      "median": 0.45316326530612244,
+      "q75": 0.4608163265306122,
+      "q90": 0.4635102040816327,
+      "max": 0.4669387755102041,
+      "mean": 0.45334693877551013
+    },
+    "duration_seconds": {
+      "min": 9.628000457305461,
+      "q10": 9.917200471041724,
+      "q25": 10.149000482051633,
+      "median": 10.454000496538356,
+      "q75": 11.031000523944385,
+      "q90": 11.649600553326309,
+      "max": 12.53200059523806,
+      "mean": 10.60952050392516
+    },
+    "frames": {
+      "min": 302.0,
+      "q10": 311.0,
+      "q25": 319.0,
+      "median": 328.0,
+      "q75": 345.75,
+      "q90": 365.1,
+      "max": 393.0,
+      "mean": 333.02
+    },
+    "placement_error_mm": {
+      "min": 0.9223316816245355,
+      "q10": 4.740607810618222,
+      "q25": 5.9057791678636145,
+      "median": 7.264740963753441,
+      "q75": 9.624124141518175,
+      "q90": 12.049785319197095,
+      "max": 13.672449866804117,
+      "mean": 7.779289416589755
+    },
+    "upright_axis_error_degrees": {
+      "min": 0.12523307506561432,
+      "q10": 0.6915871522562761,
+      "q25": 1.253221590041893,
+      "median": 1.8009010928711464,
+      "q75": 2.332611434101384,
+      "q90": 2.8785871004846046,
+      "max": 3.2803159778039106,
+      "mean": 1.7805040597025663
+    },
+    "unique_scene_seeds": 30,
+    "unique_motion_tuples": 43,
+    "unique_scene_motion_tuples": 50
+  },
   "quality_audit": {
     "audited_at_utc": "2026-07-28T15:39:36.607029+00:00",
     "repository_checker_clean_episodes": 50,
@@ -59,6 +170,29 @@
     "wrong_arm_language_labels": 0,
     "issues": []
   },
+  "training_exports_validated": {
+    "aloha_hdf5": "policy/PUMA/converted_data/aloha_hdf5/adjust_bottle-demo_dynamicvla_follow_dataset-50",
+    "lerobot_v21": "policy/PUMA/converted_data/lerobot/robotwin/adjust_bottle_follow_fast",
+    "puma_data_root": "policy/PUMA/converted_data/lerobot/robotwin",
+    "puma_data_mix": "robotwin_adjust_bottle_follow_fast",
+    "episodes": 50,
+    "training_frames": 16601,
+    "cameras": [
+      "cam_high",
+      "cam_left_wrist",
+      "cam_right_wrist"
+    ],
+    "fps": 15,
+    "puma_loader_validated": true,
+    "puma_loader_samples_validated": [
+      0,
+      8300,
+      16600
+    ],
+    "normalization_statistics_cached": true,
+    "step_index_cached": true,
+    "lerobot_videos_fully_decoded": 150
+  },
   "repository": {
     "commit": "1c215da781d38cd7defe9aac7a0c9eee661e8726",
     "dirty": true,