Organize PROWBench inputs for browsing

README.md

@@ -9,17 +9,31 @@ tags:
 - video-generation
 - programmable-world-models
 pretty_name: PROWBench — No-depth input bundle
+configs:
+- config_name: default
+  default: true
+  data_files:
+  - split: full
+    path: splits/full.parquet
+  - split: long_video
+    path: splits/long_video.parquet
+  - split: multi_view
+    path: splits/multi_view.parquet
+  - split: verified_first_frame
+    path: splits/verified_first_frame.parquet
+  - split: unverified_first_frame
+    path: splits/unverified_first_frame.parquet
+  - split: neon
+    path: splits/neon.parquet
 ---
 
 # PROWBench — No-depth input bundle
 
-Private staging snapshot `20261008T152517Z` of the complete no-depth input bundle.
+PROWBench provides structured conditioning inputs for video/world-model evaluation. This **private** dataset currently contains **170 episodes, 200 selected camera tracks, and 600 representation samples**. Each view has three proxy representations: `white`, `mixed`, and `colored-obb`. The source snapshot is `20261008T152517Z`.
 
-[Project page](https://alaya-lab.github.io/PROWBench/) · [Gallery](https://alaya-lab.github.io/PROWBench/gallery.html) · [Code repository](https://github.com/AlayaLab/PROWBench)
+[Project page](https://alaya-lab.github.io/PROWBench/) · [Gallery](https://alaya-lab.github.io/PROWBench/gallery.html) · [Project repository](https://github.com/AlayaLab/PROWBench)
 
-The archive contains **170 episodes, 200 selected camera tracks, and 600 representation samples** (`white`, `mixed`, and `colored-obb`). It preserves the existing scene assets, recorded states, camera tracks, proxy videos, prompts, available RGB first frames, bundled tools, evaluation code, and provenance. Depth is not included. Generated model result videos are separate from these conditioning inputs.
-
-| Group | Episodes | Views | Samples | Views with RGB first frame |
+| Group | Episodes | Views | Samples | Views with first frame |
 |---|---:|---:|---:|---:|
 | `long_video` | 10 | 10 | 30 | 0 |
 | `multi_view` | 10 | 40 | 120 | 0 |
@@ -27,9 +41,26 @@ The archive contains **170 episodes, 200 selected camera tracks, and 600 represe
 | `unverified_first_frame` | 90 | 90 | 270 | 0 |
 | `neon` | 20 | 20 | 60 | 20 |
 
-## Download and extract
+The videos in this repository are **proxy input videos**, not generated model results. RGB first frames are included only where recorded in the source catalog. Depth and generated result videos are not part of this release.
+
+## Browse the data
 
-Authenticate with an account that can access this private repository, then run:
+The `full` split has one row per representation sample (600 rows). The other five splits select rows from that same index by group; they are **not additional samples**. Each row includes the episode, camera, representation, prompt text, paths to directly downloadable files, source paths, SHA-256 hashes, and visual-revision indicators.
+
+```text
+splits/                   Parquet sample index: full + five group splits
+cases/                    170 episode summaries, each listing its samples
+videos/                   600 individual proxy MP4 files
+images/                   60 available RGB first-frame PNG files
+prompts/                  690 individual prompt files, including long-video windows
+archives/                 Complete, independently verified source bundle
+organized-files.jsonl.gz  Checksums and source paths for the browsable files
+organization.json         Layout and count summary
+```
+
+Paths in the Parquet columns (`video_file`, `prompt_file`, `first_frame_file`, and `case_file`) are relative to the repository root. `first_frame_file` is null when no first frame exists. For example, download one video with `hf download AlayaLab/PROWBench <video_file> --repo-type dataset`. An account with access to this private repository must be authenticated first.
+
+For the complete source bundle:
 
 ```bash
 hf download AlayaLab/PROWBench --repo-type dataset --local-dir prowbench-download
@@ -38,14 +69,12 @@ sha256sum -c SHA256SUMS
 tar --zstd -xf archives/PROWBench-no-depth-20261008T152517Z.tar.zst
 ```
 
-The extracted entry point is `inputs_no_depth/README.txt`. Keep the entire extracted directory, including `.shared`; relative symlinks and Coastal metadata references are part of the bundle. `bundle_io.py` reads the shared metadata. See `inputs_no_depth/eval/BUNDLE.md` for evaluation usage.
-
-The source snapshot stores the Gunfight records in `episodes/gunfight`, while its existing catalogs reference `episodes/neon`. This archive includes the compatibility symlink `episodes/neon -> gunfight` so those 60 sample paths resolve. The recorded data, videos, prompts, and metadata text are unchanged; `release.json` documents this packaging adjustment.
+The extracted entry point is `inputs_no_depth/README.txt`. Keep the whole directory, including `.shared`; relative symlinks and Coastal metadata references are part of the bundle. See `inputs_no_depth/eval/BUNDLE.md` for evaluation usage. The browsable folders are convenience copies of selected inputs; the archive is the complete reproducibility package.
 
-## Snapshot and reproducibility
+## Provenance and limitations
 
-`release.json` records the current counts, archive checksum, and visual revision. `snapshot-files.jsonl.gz` lists the archived paths, file hashes, and symlink targets, relative to `inputs_no_depth/`.
+`release.json` records the source snapshot, archive checksum, and visual revisions. `snapshot-files.jsonl.gz` lists the archived paths, file hashes, and symlink targets. `organized-files.jsonl.gz` separately verifies the directly browsable copies and records their source paths. `SHA256SUMS` verifies the top-level release files and split indexes.
 
-This snapshot includes the current input revisions recorded in its provenance. Historical model outputs and paper results must be associated with the input versions used to generate them. The archive has been checked against an independent local snapshot; this packaging verification does not constitute a model evaluation run.
+The source snapshot stores Gunfight records in `episodes/gunfight`, while its existing catalogs call the group `neon`. The archive includes the compatibility symlink `episodes/neon -> gunfight`, so the catalog paths resolve. The recorded content is unchanged. The Parquet field `video_matches_source` is false for **76** samples whose proxy video has a documented visual revision relative to its source video; this is a provenance distinction, not a missing file. See the archive's `provenance/` records and `promptBinding` in `samples.jsonl` before comparing historical model outputs.
 
-This is a private staging upload. Public release terms and the final dataset card remain to be prepared; existing component notices are retained inside the archive.
+No public license has been assigned to this private staging release. Component notices remain in the source bundle. This packaging check does not constitute a model evaluation run.

SHA256SUMS

@@ -1,4 +1,12 @@
-eb177ac0fa156966f0ec382de9fec5d217401a2b9246c4b31f6e20ed98ad248b  README.md
+071fc519505ca90032ca1e34672626d217ea4f16c79ac4b759ea1fdc973dd780  README.md
 c190eb114e83677c00e0527c663c99d7435c37972951806f40654d7e23a1dca3  archives/PROWBench-no-depth-20261008T152517Z.tar.zst
+91209b9a1e24da146222acd567ed7a712d0a79a12ea3bd619a383f026cddf20a  organization.json
+c27cc0fc6105d9404ab6fe601502f80c2d2682a588591acdcc9881f1ae1a47e5  organized-files.jsonl.gz
 cf5f40deb82852bb1b8b1f36e1b05d47c1e3924bd8a3664f060c7dca56c8d773  release.json
 dabb5f5cbb5d93dece6a3ea272248641fd2aaf731826b17607e99969c4ead353  snapshot-files.jsonl.gz
+d95aa0a8d0b3a13881cca979c0a391b545b19a37359063a1c9bb370997375e77  splits/full.parquet
+2e9b1cf87a5f8cafa36bcda63f849cd3a2d348c29b70ad6165276966ce6a216c  splits/long_video.parquet
+a8c66dd0dfd80ca715d7dfdbd3e43824fd9be8cf3a95bab9b33208409f88f305  splits/multi_view.parquet
+99ab9c5c7e67ab6a09597444f339307fba47e94ded2ec6f856daec02280fe84b  splits/neon.parquet
+a44503e6f8bcba1df153c0fd9338f84cf3c94eebc262634ba942beb1352a3423  splits/unverified_first_frame.parquet
+af9f1e695beb137bda1bb4a1a2ab6a575a70120343816a8e5dc4d8eb67e60cd2  splits/verified_first_frame.parquet

organization.json

@@ -0,0 +1,23 @@
+{
+  "camera_tracks": 200,
+  "direct_bytes": 486600278,
+  "direct_files": 1526,
+  "episodes": 170,
+  "first_frames": 60,
+  "layout": "browsable-copy-v1",
+  "prompt_files": 690,
+  "repo_id": "AlayaLab/PROWBench",
+  "samples": 600,
+  "source_archive_sha256": "c190eb114e83677c00e0527c663c99d7435c37972951806f40654d7e23a1dca3",
+  "source_snapshot": "20261008T152517Z",
+  "split_rows": {
+    "full": 600,
+    "long_video": 30,
+    "multi_view": 120,
+    "neon": 60,
+    "unverified_first_frame": 270,
+    "verified_first_frame": 120
+  },
+  "video_files": 600,
+  "visual_revision_samples": 76
+}

organized-files.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c27cc0fc6105d9404ab6fe601502f80c2d2682a588591acdcc9881f1ae1a47e5
+size 82259