Unify the ScaleNav data catalog and subset preparation links

README.md

@@ -24,39 +24,67 @@ Pretraining for Language-Conditioned Local Navigation**.
 [Model](https://huggingface.co/Yangyihui/ScaleNav) ·
 [Project page](https://zju3dv.github.io/ScaleNav/) · [Collection](https://huggingface.co/collections/Yangyihui/scalenav-6ad591e8ff62bb75ec75f645)
 
-## Download and prepare
-
-From the ScaleNav code repository, download the small Internet training pack:
+## Choose a dataset
+
+Choose a test subset to try navigation, or a training selection to prepare
+training videos and labels. Each download guide covers **all required sources**.
+
+| Dataset | Contents | Total download | Total data storage | Download |
+| --- | --- | ---: | ---: | --- |
+| Habitat-GS-LN test subset | 16 tasks / 13 scenes | 991.32 MB | 2.96 GB | [Download](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/EVALUATION.md#test-subset) |
+| Full Habitat-GS-LN evaluation | 157 tasks / 45 scenes | 3.17 GB | 10.36 GB | [Download](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/EVALUATION.md#full-habitat-gs-ln) |
+| Short-Horizon OVON test subset | 7 tasks / 7 scenes | 642.85 MB | 642.90 MB | [Download](#ovon-test-subset) |
+| Full Short-Horizon OVON evaluation | 524 tasks / 36 scenes | 5.84 GB | 3.67 GB | [Download](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/EVALUATION.md#full-short-horizon-ovon) |
+| Small training subset (for preview) | 17 Internet segments / 4 simulated episodes | 55.60 MB | 239.54 MB | [Download](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/DATA.md#small-training-subset) |
+| Full training data | 18,249 Internet segments / 2,600 simulated episodes, including validation | 36.10 GB | 153.04 GB | [Download](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/DATA.md#full-training-data) |
+
+Sizes use decimal MB/GB, rounded from file-byte totals. Total download counts
+files fetched from all sources; our labels are already included in the code
+repository. Total data storage includes prepared files and retained downloads.
+Full OVON assumes the documented archive cleanup; allow 9.34 GB before cleanup.
+Model weights, environments, training latent caches and run outputs are separate.
+Exact download bytes and storage breakdowns are in the
+[data guide](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/DATA.md#choose-a-download).
+
+### OVON test subset
+
+This ScaleNav subset contains 7 selected Short-Horizon OVON tasks in 7 HM3D
+scenes. Accept the Matterport academic-use agreement on the
+[subset access page](https://huggingface.co/datasets/Yangyihui/ScaleNav-OVON-Subset).
+Access is granted automatically after submitting the form. After completing
+the [installation](https://github.com/MachEmbodied/ScaleNav#installation), run
+from the ScaleNav repository root:
 
 ```bash
-python scripts/prepare_data.py internet --subset smoke --split train \
-  --source data/downloads/scalenav --output data/smoke/goto/internet \
-  --instructions goto --download
+hf auth login
+python scripts/prepare_data.py ovon --subset test_subset \
+  --source data/upstream/evaluation/ovon \
+  --scenes data/upstream/evaluation/hm3d \
+  --output data/evaluation/ovon_test_subset --download
 ```
 
-The script downloads a fixed release, verifies archive and video SHA-256 hashes,
-extracts the selected videos and writes paired annotations and training configs.
-It preserves the prepared MP4 files without another encoding pass. The default
-path downloads only ScaleNav packs; original SpatialVID archives are optional.
-
-| Selection | Samples | Download |
-| --- | ---: | --- |
-| Small Internet training subset | 17 segments / 14 clips | About 10.6 MB, plus the video index |
-| Internet training | 18,184 segments / 13,605 clips | About 8.6 GB in four independent TAR packs |
-| Internet validation | 65 segments / 42 clips | About 37 MB |
-| Annotation package | Internet/simulation labels, poses and evaluation selections | About 4.8 MB; also included in the code repository |
-
-For all Internet videos, omit `--subset smoke --split train` and use
-`--output data/training/goto/internet`. For validation only, use `--split val`
-and a separate output directory. Packs remain in `--source` for reuse; omit
-`--download` to prepare from an existing local cache. The small training subset
-is part of the training split, not a navigation benchmark.
-
-See the [data guide](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/DATA.md)
-for simulation preparation and latent caching, and the
-[training guide](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/TRAINING.md)
-for the four training phases. The combined small training subset uses the video
-pack above and one approximately 44 MB scene from the official InteriorGS release.
+The command downloads only the required scene files, original navmeshes,
+semantic labels, task metadata and official ABotN starting poses, verifies their
+SHA-256 hashes and applies our instructions. No full HM3D archives are needed.
+Continue with [OVON evaluation](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/EVALUATION.md#run-short-horizon-ovon-evaluation)
+using `EVAL_SET=ovon_test_subset`.
+
+The files are hosted in a separate access-controlled repository for the
+Matterport agreement. This page is the shared entry point for all ScaleNav data.
+
+### InteriorGS scenes
+
+InteriorGS supports individual scene and file downloads. After obtaining
+[official access](https://huggingface.co/datasets/spatialverse/InteriorGS), the
+preparation commands fetch only the files needed by the selected data split.
+The small training preview downloads one compressed scene
+(`0003_839989`, 43,943,514 bytes), alongside the 11,660,372-byte Internet pack
+and index. Habitat-GS-LN preparation likewise fetches only the required scene
+assets and labels. Neither requires downloading the entire InteriorGS dataset.
+
+Scene files remain at their official sources under the InteriorGS terms.
+Our instructions, recorded trajectories and permitted Habitat-GS navmeshes
+are supplied separately and combined automatically during preparation.
 
 ## Contents
 
@@ -87,22 +115,10 @@ Download InteriorGS scenes from the official source and replay our recorded
 poses to render training video. Simulation scenes, rendered simulation videos
 and latent caches are not included in these Internet video packs.
 
-The [evaluation guide](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/EVALUATION.md)
-provides separate preparation and evaluation commands for:
-
-| Benchmark | Tasks | Total data download |
-| --- | ---: | --- |
-| Habitat-GS-LN test subset | 16 tasks in 13 scenes | 991.32 MB |
-| Full Habitat-GS-LN | 157 tasks in 45 scenes | 3.17 GB |
-| [Short-Horizon OVON test subset](https://huggingface.co/datasets/Yangyihui/ScaleNav-OVON-Subset) | 7 tasks in 7 scenes | 642.85 MB |
-| Full Short-Horizon OVON | 524 tasks in 36 scenes | 5.84 GB |
-
-The curated test subsets list paper examples first; use the full sets for benchmark scores.
-The OVON test subset downloads only its required scene and task files from a
-[dedicated repository](https://huggingface.co/datasets/Yangyihui/ScaleNav-OVON-Subset),
-after accepting the Matterport agreement. Full OVON uses the official archives.
-Download sizes include all required data sources; prepared scene sizes and
-commands are in the evaluation guide.
+The [evaluation guide](https://github.com/MachEmbodied/ScaleNav/blob/main/docs/EVALUATION.md) provides preparation and
+evaluation commands for both benchmarks. The test subsets contain selected
+successful ScaleNav cases, with paper examples first; use the full sets for
+benchmark scores.
 
 The preparation command downloads the required original Habitat-GS navmeshes
 from this repository and verifies their SHA-256 hashes. Full Habitat-GS-LN