📚 Document the processed training assets

Add the canonical 50-scene layout, download command, reset-sampling semantics, and current asset availability to the dataset card.

README.md

@@ -11,9 +11,9 @@ tags:
 
 # 🧭 NavOL artifacts
 
-This dataset repository contains the checkpoints and benchmark data used by
-NavOL. Source code and executable data tools are maintained at
-<https://github.com/WAboutMe/NavOL>.
+This dataset repository contains the checkpoints, processed 50-scene training
+asset, and benchmark data used by NavOL. Source code and executable data tools
+are maintained at <https://github.com/WAboutMe/NavOL>.
 
 ## 🧠 Checkpoints
 
@@ -27,6 +27,32 @@ NavOL. Source code and executable data tools are maintained at
 The filenames retain their training mode and iteration and should not be
 treated as interchangeable copies of one generic checkpoint.
 
+## 🏋️ Processed training data
+
+The canonical processed training asset is stored at:
+
+```text
+datasets/train/3d_front_scene_50/
+├── index.json
+├── selected.json
+├── scene.glb
+├── usd/
+│   ├── config.yaml
+│   ├── scene.usd
+│   └── textures/
+└── navmesh_scenes/scene_*.glb
+```
+
+`selected.json` records the 50-scene selection. The merged `scene.usd` and its
+textures are loaded by Isaac Sim, while the 50 per-scene GLBs under
+`navmesh_scenes/` are used by Habitat-Sim planning. `scene.glb` is retained as
+a portable reconstruction source, and `usd/config.yaml` is an informational,
+path-sanitized conversion record rather than a training input.
+
+The training asset intentionally does not include `sample_100.npy`. Canonical
+random training uses `sample_from_npy=False`; a fixed reset array is needed
+only when that option is explicitly enabled.
+
 ## 🗺️ Benchmark data
 
 | Split | Processed archive | Raw archive |
@@ -54,6 +80,15 @@ hf download WAboutme/NavOL \
   --local-dir assets
 ```
 
+Download the processed training asset into a NavOL asset root:
+
+```bash
+hf download WAboutme/NavOL \
+  --repo-type dataset \
+  --include "datasets/train/3d_front_scene_50/**" \
+  --local-dir assets
+```
+
 Download the benchmark archives:
 
 ```bash
@@ -70,10 +105,22 @@ NavOL resolves the default checkpoint at:
 assets/models/checkpoints/navol-mpc-iter1000.pt
 ```
 
+The training download resolves to:
+
+```text
+assets/datasets/train/3d_front_scene_50/
+```
+
+This repository does **not** currently distribute the NavDP initialization
+checkpoint `models/navdp-cross-modal.ckpt` or the robot asset
+`robots/dingo.usd`. Obtain both under their respective upstream terms and
+place them in the documented asset layout before training.
+
 ## 📄 License
 
 NavOL-authored source and configuration files use the BSD 3-Clause License.
-Checkpoints, benchmark scenes, textures, and other third-party-derived
-artifacts retain the terms of their respective upstream sources. The dataset
+Checkpoints, training and benchmark scenes, textures, and other
+third-party-derived artifacts retain the terms of their respective upstream
+sources. The dataset
 repository therefore uses `license: other`. See [LICENSE.md](LICENSE.md) and
 [THIRD_PARTY_NOTICES.md](THIRD_PARTY_NOTICES.md).