Guan123 commited on
Commit
2fbef61
·
verified ·
1 Parent(s): 73a889e

Add dataset card, license, annotations, splits, bench and video index

Browse files
.gitattributes CHANGED
@@ -58,3 +58,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
58
  # Video files - compressed
59
  *.mp4 filter=lfs diff=lfs merge=lfs -text
60
  *.webm filter=lfs diff=lfs merge=lfs -text
 
 
58
  # Video files - compressed
59
  *.mp4 filter=lfs diff=lfs merge=lfs -text
60
  *.webm filter=lfs diff=lfs merge=lfs -text
61
+ annotations/annotations_80k.jsonl filter=lfs diff=lfs merge=lfs -text
LICENSE ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ Apache License
3
+ Version 2.0, January 2004
4
+ http://www.apache.org/licenses/
5
+
6
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
7
+
8
+ 1. Definitions.
9
+
10
+ "License" shall mean the terms and conditions for use, reproduction,
11
+ and distribution as defined by Sections 1 through 9 of this document.
12
+
13
+ "Licensor" shall mean the copyright owner or entity authorized by
14
+ the copyright owner that is granting the License.
15
+
16
+ "Legal Entity" shall mean the union of the acting entity and all
17
+ other entities that control, are controlled by, or are under common
18
+ control with that entity. For the purposes of this definition,
19
+ "control" means (i) the power, direct or indirect, to cause the
20
+ direction or management of such entity, whether by contract or
21
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
22
+ outstanding shares, or (iii) beneficial ownership of such entity.
23
+
24
+ "You" (or "Your") shall mean an individual or Legal Entity
25
+ exercising permissions granted by this License.
26
+
27
+ "Source" form shall mean the preferred form for making modifications,
28
+ including but not limited to software source code, documentation
29
+ source, and configuration files.
30
+
31
+ "Object" form shall mean any form resulting from mechanical
32
+ transformation or translation of a Source form, including but
33
+ not limited to compiled object code, generated documentation,
34
+ and conversions to other media types.
35
+
36
+ "Work" shall mean the work of authorship, whether in Source or
37
+ Object form, made available under the License, as indicated by a
38
+ copyright notice that is included in or attached to the work
39
+ (an example is provided in the Appendix below).
40
+
41
+ "Derivative Works" shall mean any work, whether in Source or Object
42
+ form, that is based on (or derived from) the Work and for which the
43
+ editorial revisions, annotations, elaborations, or other modifications
44
+ represent, as a whole, an original work of authorship. For the purposes
45
+ of this License, Derivative Works shall not include works that remain
46
+ separable from, or merely link (or bind by name) to the interfaces of,
47
+ the Work and Derivative Works thereof.
48
+
49
+ "Contribution" shall mean any work of authorship, including
50
+ the original version of the Work and any modifications or additions
51
+ to that Work or Derivative Works thereof, that is intentionally
52
+ submitted to Licensor for inclusion in the Work by the copyright owner
53
+ or by an individual or Legal Entity authorized to submit on behalf of
54
+ the copyright owner. For the purposes of this definition, "submitted"
55
+ means any form of electronic, verbal, or written communication sent
56
+ to the Licensor or its representatives, including but not limited to
57
+ communication on electronic mailing lists, source code control systems,
58
+ and issue tracking systems that are managed by, or on behalf of, the
59
+ Licensor for the purpose of discussing and improving the Work, but
60
+ excluding communication that is conspicuously marked or otherwise
61
+ designated in writing by the copyright owner as "Not a Contribution."
62
+
63
+ "Contributor" shall mean Licensor and any individual or Legal Entity
64
+ on behalf of whom a Contribution has been received by Licensor and
65
+ subsequently incorporated within the Work.
66
+
67
+ 2. Grant of Copyright License. Subject to the terms and conditions of
68
+ this License, each Contributor hereby grants to You a perpetual,
69
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
70
+ copyright license to reproduce, prepare Derivative Works of,
71
+ publicly display, publicly perform, sublicense, and distribute the
72
+ Work and such Derivative Works in Source or Object form.
73
+
74
+ 3. Grant of Patent License. Subject to the terms and conditions of
75
+ this License, each Contributor hereby grants to You a perpetual,
76
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
77
+ (except as stated in this section) patent license to make, have made,
78
+ use, offer to sell, sell, import, and otherwise transfer the Work,
79
+ where such license applies only to those patent claims licensable
80
+ by such Contributor that are necessarily infringed by their
81
+ Contribution(s) alone or by combination of their Contribution(s)
82
+ with the Work to which such Contribution(s) was submitted. If You
83
+ institute patent litigation against any entity (including a
84
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
85
+ or a Contribution incorporated within the Work constitutes direct
86
+ or contributory patent infringement, then any patent licenses
87
+ granted to You under this License for that Work shall terminate
88
+ as of the date such litigation is filed.
89
+
90
+ 4. Redistribution. You may reproduce and distribute copies of the
91
+ Work or Derivative Works thereof in any medium, with or without
92
+ modifications, and in Source or Object form, provided that You
93
+ meet the following conditions:
94
+
95
+ (a) You must give any other recipients of the Work or
96
+ Derivative Works a copy of this License; and
97
+
98
+ (b) You must cause any modified files to carry prominent notices
99
+ stating that You changed the files; and
100
+
101
+ (c) You must retain, in the Source form of any Derivative Works
102
+ that You distribute, all copyright, patent, trademark, and
103
+ attribution notices from the Source form of the Work,
104
+ excluding those notices that do not pertain to any part of
105
+ the Derivative Works; and
106
+
107
+ (d) If the Work includes a "NOTICE" text file as part of its
108
+ distribution, then any Derivative Works that You distribute must
109
+ include a readable copy of the attribution notices contained
110
+ within such NOTICE file, excluding those notices that do not
111
+ pertain to any part of the Derivative Works, in at least one
112
+ of the following places: within a NOTICE text file distributed
113
+ as part of the Derivative Works; within the Source form or
114
+ documentation, if provided along with the Derivative Works; or,
115
+ within a display generated by the Derivative Works, if and
116
+ wherever such third-party notices normally appear. The contents
117
+ of the NOTICE file are for informational purposes only and
118
+ do not modify the License. You may add Your own attribution
119
+ notices within Derivative Works that You distribute, alongside
120
+ or as an addendum to the NOTICE text from the Work, provided
121
+ that such additional attribution notices cannot be construed
122
+ as modifying the License.
123
+
124
+ You may add Your own copyright statement to Your modifications and
125
+ may provide additional or different license terms and conditions
126
+ for use, reproduction, or distribution of Your modifications, or
127
+ for any such Derivative Works as a whole, provided Your use,
128
+ reproduction, and distribution of the Work otherwise complies with
129
+ the conditions stated in this License.
130
+
131
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
132
+ any Contribution intentionally submitted for inclusion in the Work
133
+ by You to the Licensor shall be under the terms and conditions of
134
+ this License, without any additional terms or conditions.
135
+ Notwithstanding the above, nothing herein shall supersede or modify
136
+ the terms of any separate license agreement you may have executed
137
+ with Licensor regarding such Contributions.
138
+
139
+ 6. Trademarks. This License does not grant permission to use the trade
140
+ names, trademarks, service marks, or product names of the Licensor,
141
+ except as required for reasonable and customary use in describing the
142
+ origin of the Work and reproducing the content of the NOTICE file.
143
+
144
+ 7. Disclaimer of Warranty. Unless required by applicable law or
145
+ agreed to in writing, Licensor provides the Work (and each
146
+ Contributor provides its Contributions) on an "AS IS" BASIS,
147
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
148
+ implied, including, without limitation, any warranties or conditions
149
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
150
+ PARTICULAR PURPOSE. You are solely responsible for determining the
151
+ appropriateness of using or redistributing the Work and assume any
152
+ risks associated with Your exercise of permissions under this License.
153
+
154
+ 8. Limitation of Liability. In no event and under no legal theory,
155
+ whether in tort (including negligence), contract, or otherwise,
156
+ unless required by applicable law (such as deliberate and grossly
157
+ negligent acts) or agreed to in writing, shall any Contributor be
158
+ liable to You for damages, including any direct, indirect, special,
159
+ incidental, or consequential damages of any character arising as a
160
+ result of this License or out of the use or inability to use the
161
+ Work (including but not limited to damages for loss of goodwill,
162
+ work stoppage, computer failure or malfunction, or any and all
163
+ other commercial damages or losses), even if such Contributor
164
+ has been advised of the possibility of such damages.
165
+
166
+ 9. Accepting Warranty or Additional Liability. While redistributing
167
+ the Work or Derivative Works thereof, You may choose to offer,
168
+ and charge a fee for, acceptance of support, warranty, indemnity,
169
+ or other liability obligations and/or rights consistent with this
170
+ License. However, in accepting such obligations, You may act only
171
+ on Your own behalf and on Your sole responsibility, not on behalf
172
+ of any other Contributor, and only if You agree to indemnify,
173
+ defend, and hold each Contributor harmless for any liability
174
+ incurred by, or claims asserted against, such Contributor by reason
175
+ of your accepting any such warranty or additional liability.
176
+
177
+ END OF TERMS AND CONDITIONS
178
+
179
+ APPENDIX: How to apply the Apache License to your work.
180
+
181
+ To apply the Apache License to your work, attach the following
182
+ boilerplate notice, with the fields enclosed by brackets "[]"
183
+ replaced with your own identifying information. (Don't include
184
+ the brackets!) The text should be enclosed in the appropriate
185
+ comment syntax for the file format. We also recommend that a
186
+ file or class name and description of purpose be included on the
187
+ same "printed page" as the copyright notice for easier
188
+ identification within third-party archives.
189
+
190
+ Copyright [yyyy] [name of copyright owner]
191
+
192
+ Licensed under the Apache License, Version 2.0 (the "License");
193
+ you may not use this file except in compliance with the License.
194
+ You may obtain a copy of the License at
195
+
196
+ http://www.apache.org/licenses/LICENSE-2.0
197
+
198
+ Unless required by applicable law or agreed to in writing, software
199
+ distributed under the License is distributed on an "AS IS" BASIS,
200
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
201
+ See the License for the specific language governing permissions and
202
+ limitations under the License.
README.md CHANGED
@@ -1,3 +1,194 @@
1
  ---
2
  license: apache-2.0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
+ pretty_name: SyncReward-Data-80K
4
+ task_categories:
5
+ - video-classification
6
+ tags:
7
+ - audio-video
8
+ - audio-visual
9
+ - synchronization
10
+ - reward-model
11
+ - human-feedback
12
+ - video-generation
13
+ size_categories:
14
+ - 10K<n<100K
15
+ configs:
16
+ - config_name: annotations
17
+ default: true
18
+ data_files:
19
+ - split: all
20
+ path: annotations/annotations_80k.jsonl
21
+ - config_name: stage2
22
+ data_files:
23
+ - split: train
24
+ path: splits/stage2_train.jsonl
25
+ - split: validation
26
+ path: splits/stage2_val.jsonl
27
+ - config_name: bench
28
+ data_files:
29
+ - split: test
30
+ path: bench/syncreward_bench.jsonl
31
  ---
32
+
33
+ # SyncReward-Data-80K
34
+
35
+ This dataset contains:
36
+
37
+ - human audio-visual **synchronization ratings** of 57,099 generated audio-video clips;
38
+ - the clips themselves, as mp4 files in zip archives;
39
+ - the Stage-2 training splits derived from the ratings;
40
+ - **SyncReward-Bench**, a 1,000-clip benchmark for sync metrics.
41
+
42
+ It is the data behind the [SyncReward](https://huggingface.co/Guan123/SyncReward) reward model.
43
+
44
+ Paper, code repository and citation: coming soon.
45
+
46
+ ## Repository layout
47
+
48
+ ```
49
+ annotations/annotations_80k.jsonl 80,001 ratings (one row per annotator x clip)
50
+ splits/stage2_train.jsonl 54,463 clips, Stage-2 training targets
51
+ splits/stage2_val.jsonl 1,636 clips, Stage-2 validation targets
52
+ bench/syncreward_bench.jsonl 1,000 clips, SyncReward-Bench
53
+ videos/<model>_partNNN.zip 57,099 mp4 files in 19 zip archives (213 GB)
54
+ videos/index.jsonl video_path -> zip archive
55
+ scripts/rebuild_targets.py rebuilds splits/ from annotations/
56
+ SHA256SUMS checksums of all files except README.md
57
+ ```
58
+
59
+ ## Download
60
+
61
+ ```bash
62
+ hf download Guan123/SyncReward-Data --repo-type dataset --local-dir SyncReward-Data
63
+ cd SyncReward-Data
64
+ sha256sum -c SHA256SUMS
65
+ for z in videos/*.zip; do unzip -q -n "$z"; done # -> generated/<model>/*.mp4
66
+ ```
67
+
68
+ Every `video_path` in the jsonl files is relative to the extraction directory. Pass that directory as
69
+ `--video_root` to the SyncReward code. To fetch only the metadata, or only one generator, use for example
70
+ `--include 'annotations/*' --include 'splits/*' --include 'bench/*'` or `--include 'videos/kling_*'`.
71
+ `videos/index.jsonl` tells you which archive holds a given clip.
72
+
73
+ ```python
74
+ from datasets import load_dataset
75
+ ratings = load_dataset("Guan123/SyncReward-Data", "annotations", split="all")
76
+ stage2 = load_dataset("Guan123/SyncReward-Data", "stage2") # train / validation
77
+ bench = load_dataset("Guan123/SyncReward-Data", "bench", split="test")
78
+ ```
79
+
80
+ ## Videos
81
+
82
+ | Generator (`model`) | Videos | Size | Archives |
83
+ |---|---|---|---|
84
+ | `davinci` | 6,691 | 7.95 GB | 1 |
85
+ | `kling` | 6,718 | 47.43 GB | 4 |
86
+ | `ltx2.3` | 6,772 | 18.23 GB | 2 |
87
+ | `mova` | 6,661 | 30.42 GB | 3 |
88
+ | `nava` | 6,796 | 23.17 GB | 2 |
89
+ | `ovi` | 6,733 | 7.85 GB | 1 |
90
+ | `seedance` | 6,728 | 24.30 GB | 2 |
91
+ | `veo3` | 10,000 | 53.87 GB | 4 |
92
+ | **Total** | **57,099** | **213.2 GB** | **19** |
93
+
94
+ The archives are uncompressed (stored) zips; mp4 is already compressed. Every clip referenced by any
95
+ jsonl file is included. The clips were generated from text prompts by Kling 3.0, Seedance 2.0, Veo 3,
96
+ LTX-2.3, daVinci, MOVA, NAVA and Ovi. Each generator's own terms of use also apply to its outputs (see
97
+ [License](#license)).
98
+
99
+ ## Training data
100
+
101
+ **SyncReward-Data-80K** contains 80,001 human synchronization ratings from 5 annotators over 57,099 clips
102
+ generated by 8 text-to-audio-video models (about 10k ratings per model).
103
+
104
+ Each rating first marks whether the audio is **related** to the visible scene at all. If it is, the
105
+ annotator scores synchronization as 0 (out of sync), 1 (partially synchronized) or 2 (well synchronized).
106
+ We keep unrelated audio in the data on purpose: it is a common failure mode of current generators, and a
107
+ sync reward should penalize it rather than never see it.
108
+
109
+ | Rating | Count | Share |
110
+ |---|---|---|
111
+ | Unrelated audio | 10,065 | 12.6% |
112
+ | 0 – out of sync | 12,907 | 16.1% |
113
+ | 1 – partial | 37,935 | 47.4% |
114
+ | 2 – well synchronized | 19,094 | 23.9% |
115
+
116
+ The Stage-2 target of a clip is the rater-bias-corrected mean of its related ratings. Clips that every
117
+ rater marked unrelated (6,261 train and 172 validation clips, 11.5% of the training pool) get target 0,
118
+ the same as a clear sync failure. The 1,931 clips with mixed votes use their related ratings only.
119
+
120
+ Splits are clip-disjoint by (prompt, generator): 54,463 training and 1,636 validation clips.
121
+ The 1,000 SyncReward-Bench clips are held out from training.
122
+
123
+ ## Fields
124
+
125
+ ### `annotations`: one row per (annotator, clip)
126
+
127
+ | Field | Meaning |
128
+ |---|---|
129
+ | `uid` | prompt id; the same prompt is rendered by several generators |
130
+ | `model` | generator, see [Videos](#videos) |
131
+ | `video_path` | clip path relative to the extraction directory |
132
+ | `rater` | anonymized annotator, `rater_1` ... `rater_5` |
133
+ | `related` | 1 if the audio is related to the visible scene, 0 if not |
134
+ | `score` | 0 / 1 / 2 sync score; `null` when `related = 0` |
135
+ | `split` | the clip's split: `train`, `val` or `test` (held out) |
136
+ | `in_bench` | the clip is in SyncReward-Bench |
137
+
138
+ - **By split:** `train` has 54,463 clips and 75,609 ratings, `val` has 1,636 clips and 2,604 ratings,
139
+ and `test` has 1,000 held-out clips and 1,788 ratings.
140
+ - **Ratings per clip:** 1 (38,695 clips), 2 (14,379), 3 (3,578), 4 (421), 5 (26).
141
+ - **`in_bench`:** marks 1,880 ratings. These are the original in-pool ratings; the bench target comes
142
+ from an independent re-annotation.
143
+
144
+ ### `stage2`: one row per clip
145
+
146
+ | Field | Meaning |
147
+ |---|---|
148
+ | `video_path`, `uid`, `model` | as above |
149
+ | `target` | Stage-2 regression target in [0, 2] |
150
+ | `related` | 0 if every rater marked the audio unrelated (then `target = 0`), else 1 |
151
+
152
+ ### `bench`: one row per clip
153
+
154
+ | Field | Meaning |
155
+ |---|---|
156
+ | `id` | bench id, `curated1k_0000` ... `curated1k_0999` |
157
+ | `video_path`, `uid`, `model` | as above |
158
+ | `target` | mean of up to 3 independent human scores (0/1/2) |
159
+ | `n_labels` | number of scores averaged: 3 for 933 clips, 2 for 54, 1 for 13 |
160
+ | `prompt` | text prompt used to generate the clip |
161
+
162
+ SyncReward-Bench covers 7 generators (no Veo 3), with 142 or 143 clips each.
163
+
164
+ ## Rebuilding the Stage-2 targets
165
+
166
+ `python scripts/rebuild_targets.py --check` recomputes `splits/stage2_*.jsonl` from
167
+ `annotations/annotations_80k.jsonl` and confirms the result is byte-identical to the released files.
168
+
169
+ 1. Rater bias `b_j` = mean of (rater score - clip mean score) over clips with at least 2 scores:
170
+ rater_1 -0.042, rater_2 -0.071, rater_3 -0.231, rater_4 +0.124, rater_5 +0.223.
171
+ 2. Clip target = mean over the clip's related ratings of `clip(score - b_j, 0, 2)`, rounded to 4 decimals.
172
+ 3. Clips with no related rating get `target = 0` and `related = 0`.
173
+
174
+ ## Splits and overlap
175
+
176
+ - **Disjointness:** splits are disjoint by `(uid, model)`. The same prompt rendered by a different
177
+ generator can still appear in several splits.
178
+ - **Held-out clips:** the 1,000 `test` clips were held out from training.
179
+ - **Bench sources:** of the 1,000 SyncReward-Bench clips, 817 come from `test` and **183 from `val`**.
180
+ The validation split was never used for gradient updates, but it did select the released checkpoint, so
181
+ those 183 bench clips influenced model selection.
182
+ - **Unused test clips:** the other 183 `test` clips appear only in `annotations`.
183
+ - **Training overlap:** no bench clip is in `stage2/train`.
184
+
185
+ ## License
186
+
187
+ The annotations, manifests and scripts are released under Apache-2.0 (see [LICENSE](LICENSE)). The videos
188
+ were produced by the third-party generators listed above. By using them you also agree to follow each
189
+ generator's terms of use. They are intended for research on audio-visual synchronization and video
190
+ generation evaluation.
191
+
192
+ ## Citation
193
+
194
+ Coming soon.
SHA256SUMS ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ cfc7749b96f63bd31c3c42b5c471bf756814053e847c10f3eb003417bc523d30 LICENSE
2
+ 7253908d4926e345384272aac67d81322ab5a6b64e3eeefdbfc803a7bbe49229 annotations/annotations_80k.jsonl
3
+ 9ed8b8f38d7ac3711aaa8911f753aa558a042a4f72e3a8a461c066a105e67993 bench/syncreward_bench.jsonl
4
+ 4d131da3c7108dde156c9d9c5afb625bed64e989961eb61df38f78f496068a74 scripts/rebuild_targets.py
5
+ 83110d69b536012a759a5998a34d0c9aa0b9f03b15958923e68bf6985ca81e0e splits/stage2_train.jsonl
6
+ 0bb59d9c5a78475a939c00be69586946db04a2ea9bd5ae411571828c8f1c4149 splits/stage2_val.jsonl
7
+ 929f2714e7828beb97166bb9acc00c122d9da94f4b1605339fb79acdc3ad60e3 videos/davinci_part000.zip
8
+ adcb714b61a1c14722878d2ce4bf8590170568cdd083009fbdcb01ca6024fb30 videos/index.jsonl
9
+ 3ddb07b77a3cc743d34c4c1eafec075cb06494695733d98911aab78e989330a9 videos/kling_part000.zip
10
+ 5d2a26d79230fbfb970c486b4a2635c9f4072e01300f35d42377bffca9127035 videos/kling_part001.zip
11
+ fa74f908e350f80de9be5d16c0aa628cbeb692db99a59d611959d2ea49860dfa videos/kling_part002.zip
12
+ b7b162ab9261e0dec130d7b274d1b166781277886b0f04a9fd3709961e252aed videos/kling_part003.zip
13
+ b1743a427ce0ef099908ce646b041e7e871427979257b4b3e17f088290b817c6 videos/ltx2.3_part000.zip
14
+ 28bcf25317eee167600e8df32d001d5ddaaacd50617f4ede041cf6dab1abfaba videos/ltx2.3_part001.zip
15
+ a94c9c1123e1fdd87bc8ffc206f27d450f7bb40b3ff6f76d44ce866312e5e49a videos/mova_part000.zip
16
+ 5bb565c7f2e3824b9d791734104f1e730d700ab9af6ca34c245df15cec045d27 videos/mova_part001.zip
17
+ 29be70837f29d670b8a5a501d09306287d264ba2d4f40724ce4dd0b26c052d53 videos/mova_part002.zip
18
+ 7a61a0d060acc305cde93ea69a37a09c3b89828dae3d84c369a0d2b024aee569 videos/nava_part000.zip
19
+ b12fac10a23f2aa11804bfd00884bbf7e9f5566ff82b138672c1a58cb4f0b3b1 videos/nava_part001.zip
20
+ 382ef2663f7d93abaa8fd42466d1e007a2b190e344ba9efb8bf8469e98ba5a04 videos/ovi_part000.zip
21
+ 2343d75ef66009d2f6e5530b322a3b64587611fdf65ea10fbf0df9d78ee9f50b videos/seedance_part000.zip
22
+ 8bc6e304e945ec856a0a3d8fc1495f0c569934d9bd24c261b3e0a41ce9fdef45 videos/seedance_part001.zip
23
+ b287654686182862af1bc5bcd5c03c9ce0170ab1eda245f1936a584c3804b743 videos/veo3_part000.zip
24
+ 9d83983c1687e90f7b94c1069c9ba55822c8a287bdbb09a69151ca4f22302823 videos/veo3_part001.zip
25
+ 25d43c6444efd279e3fdb9e9fc61a326204d92a3d32ddd03d7de8adc92f19e97 videos/veo3_part002.zip
26
+ 8349f5df6d3e71ec760b67aae3a692a3bccee7db997c9ae7493c470b5bfaea21 videos/veo3_part003.zip
annotations/annotations_80k.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7253908d4926e345384272aac67d81322ab5a6b64e3eeefdbfc803a7bbe49229
3
+ size 15581161
bench/syncreward_bench.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
scripts/rebuild_targets.py ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Rebuild the Stage-2 clip targets (splits/stage2_{train,val}.jsonl) from the raw ratings.
2
+
3
+ python scripts/rebuild_targets.py # write to rebuilt/
4
+ python scripts/rebuild_targets.py --check # compare with splits/ and exit non-zero on mismatch
5
+
6
+ Target of a clip = mean over its related ratings of clip(score - rater_bias, 0, 2), rounded to 4 decimals.
7
+ rater_bias = mean of (score - clip mean score) over clips with at least 2 scores. Clips that every rater
8
+ marked unrelated get target 0 and related = 0. Rows are ordered as in the release: scored clips, then
9
+ all-unrelated clips, each sorted by (model, uid).
10
+ """
11
+ import argparse
12
+ import json
13
+ from collections import defaultdict
14
+ from pathlib import Path
15
+
16
+ ROOT = Path(__file__).resolve().parent.parent
17
+
18
+
19
+ def read_jsonl(path):
20
+ with open(path, encoding="utf-8") as f:
21
+ return [json.loads(line) for line in f if line.strip()]
22
+
23
+
24
+ def rater_bias(clips):
25
+ dev = defaultdict(list)
26
+ for ratings in clips.values():
27
+ scored = [(r["rater"], r["score"]) for r in ratings if r["score"] is not None]
28
+ if len(scored) < 2:
29
+ continue
30
+ mean = sum(s for _, s in scored) / len(scored)
31
+ for rater, s in scored:
32
+ dev[rater].append(s - mean)
33
+ return {rater: sum(v) / len(v) for rater, v in dev.items()}
34
+
35
+
36
+ def build(annotations):
37
+ clips = defaultdict(list)
38
+ for r in annotations:
39
+ clips[(r["model"], r["uid"])].append(r)
40
+ bias = rater_bias(clips)
41
+ out = {}
42
+ for split in ("train", "val"):
43
+ scored, unrelated = [], []
44
+ for key in sorted(k for k, v in clips.items() if v[0]["split"] == split):
45
+ ratings = clips[key]
46
+ cal = [min(2.0, max(0.0, r["score"] - bias[r["rater"]])) for r in ratings if r["score"] is not None]
47
+ row = {"video_path": ratings[0]["video_path"], "uid": key[1], "model": key[0]}
48
+ if cal:
49
+ scored.append({**row, "target": float(round(sum(cal) / len(cal), 4)), "related": 1})
50
+ else:
51
+ unrelated.append({**row, "target": 0.0, "related": 0})
52
+ out[split] = scored + unrelated
53
+ return out, bias
54
+
55
+
56
+ def dumps(rows):
57
+ return "".join(json.dumps(r, ensure_ascii=False) + "\n" for r in rows)
58
+
59
+
60
+ def main():
61
+ parser = argparse.ArgumentParser(description=__doc__.split("\n")[0])
62
+ parser.add_argument("--annotations", default=ROOT / "annotations" / "annotations_80k.jsonl")
63
+ parser.add_argument("--splits", default=ROOT / "splits")
64
+ parser.add_argument("--output", default=ROOT / "rebuilt")
65
+ parser.add_argument("--check", action="store_true")
66
+ args = parser.parse_args()
67
+
68
+ out, bias = build(read_jsonl(args.annotations))
69
+ print("rater bias:", {k: round(v, 4) for k, v in sorted(bias.items())})
70
+ ok = True
71
+ for split, rows in out.items():
72
+ name = f"stage2_{split}.jsonl"
73
+ if args.check:
74
+ released = (Path(args.splits) / name).read_text(encoding="utf-8")
75
+ same = released == dumps(rows)
76
+ ok &= same
77
+ print(f"{name}: {len(rows)} rows, {'identical' if same else 'DIFFERENT'} to the release")
78
+ else:
79
+ Path(args.output).mkdir(parents=True, exist_ok=True)
80
+ (Path(args.output) / name).write_text(dumps(rows), encoding="utf-8")
81
+ print(f"wrote {Path(args.output) / name} ({len(rows)} rows)")
82
+ raise SystemExit(0 if ok else 1)
83
+
84
+
85
+ if __name__ == "__main__":
86
+ main()
splits/stage2_train.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
splits/stage2_val.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
videos/index.jsonl ADDED
The diff for this file is too large to render. See raw diff