Hugging Face's logo

Datasets:
cy0307
/
ropedia-xperience-10m-task-suite-artifacts

Modalities:
Tabular
Text
Formats:
parquet
Size:
10K - 100K
Tags:
embodied-ai
multimodal
xperience-10m
evaluation
qwen3-omni
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
cy0307 commited on
Commit
120333b
·
verified ·
1 Parent(s): 3618a80

Publish Ropedia Xperience-10M derived artifacts

Browse files
Files changed (30) hide show
  1. data/artifact_index.json +326 -51
  2. data/foundation_model_plan.json +11 -6
  3. data/live_publication_status.json +749 -183
  4. data/mirror_parity.json +0 -0
  5. data/omni_finetune_verified_result.json +92 -80
  6. data/omni_model_comparison.json +523 -20
  7. data/project_packet.json +106 -89
  8. data/project_status.json +336 -296
  9. data/public_surface_qa.json +27 -26
  10. data/publication_audit.json +15 -10
  11. data/quality_gates.json +2 -2
  12. data/qwen3_full_parameter_gates.json +219 -0
  13. data/research_roadmap.json +237 -213
  14. data/research_roadmap_interactive.json +34 -11
  15. data/scope_claims_audit.json +9 -9
  16. data/source_alignment_audit.json +1 -1
  17. data/summary_metrics.json +1 -1
  18. data/task_suite_enhancement_128.json +599 -0
  19. data/task_surface_integrity.json +145 -145
  20. data/website_integrity.json +39 -28
  21. docs/data/mirror_parity.json +0 -0
  22. docs/data/public_surface_qa.json +7 -7
  23. docs/data/publication_audit.json +5 -5
  24. docs/data/quality_gates.json +1 -1
  25. docs/data/scope_claims_audit.json +1 -1
  26. docs/data/source_alignment_audit.json +1 -1
  27. docs/data/task_surface_integrity.json +144 -144
  28. docs/data/website_integrity.json +3 -3
  29. scripts/sync_hf_publish_mirrors.py +23 -0
  30. scripts/validate_mirror_parity.py +3 -0
data/artifact_index.json CHANGED
@@ -1,19 +1,19 @@
1
  {
2
  "title": "Ropedia Xperience-10M Task Suite Artifact Index",
3
- "generated_at_utc": "2026-06-07T15:47:31+00:00",
4
  "status": "pass",
5
- "artifact_count": 129,
6
  "missing": [],
7
  "by_kind": {
8
  "project_path": 14,
9
  "scaleup_contract": 7,
10
- "scaleup_status": 20,
11
  "publication_workflow": 5,
12
  "project_scope": 1,
13
  "source_alignment": 5,
14
  "evaluation_protocol": 3,
15
  "result_interpretation": 5,
16
- "metrics_source": 18,
17
  "website_data": 3,
18
  "visual_evidence": 7,
19
  "quality_gate": 12,
@@ -31,8 +31,8 @@
31
  "generated_figure_assets": 1,
32
  "citation": 1,
33
  "license": 1,
34
- "verified_public_package": 6,
35
- "publication_audit": 4
36
  },
37
  "artifacts": [
38
  {
@@ -65,8 +65,8 @@
65
  "surface": "repo_hf",
66
  "shows": "Gives a compact current-state table for first-pass readers.",
67
  "exists": true,
68
- "bytes": 9926,
69
- "sha256": "c7dfb7a45f0c1ea435c16d93208a82da4227336e34f56a96d4afa04fce42438c"
70
  },
71
  {
72
  "id": "project_status_json",
@@ -76,8 +76,8 @@
76
  "surface": "website_hf",
77
  "shows": "Machine-readable copy of the current project status for website and HF mirrors.",
78
  "exists": true,
79
- "bytes": 16455,
80
- "sha256": "3590ee1e09ecf819080a7714ea9629db305e1fd68c99a65f62bb65061c0d766c"
81
  },
82
  {
83
  "id": "research_roadmap",
@@ -87,8 +87,8 @@
87
  "surface": "repo_hf",
88
  "shows": "Defines the path from public-sample task development to multi-episode held-out evaluation and larger omni-model extensions.",
89
  "exists": true,
90
- "bytes": 12233,
91
- "sha256": "020512aa647cef7d63eccf7bb8dd6cb86f0e5c457f3c0e3d5ef293e7b35a58bf"
92
  },
93
  {
94
  "id": "research_roadmap_json",
@@ -98,8 +98,8 @@
98
  "surface": "website_hf",
99
  "shows": "Machine-readable research roadmap for the website and Hugging Face mirrors.",
100
  "exists": true,
101
- "bytes": 10133,
102
- "sha256": "45fd3a1bde93654ccfe14f9271928a67b36eb3f166826bfbdbb9c1092ad33bcf"
103
  },
104
  {
105
  "id": "foundation_model_plan",
@@ -109,8 +109,8 @@
109
  "surface": "repo_hf",
110
  "shows": "Defines the post-data-gate backbone choices: Qwen3-Omni first, Cosmos 3 for world modeling, and VLA/policy models after action-target conversion.",
111
  "exists": true,
112
- "bytes": 9075,
113
- "sha256": "444d13ab556d2e16a199a7fca191b87c85ab8685d167aab357bc6341839299a2"
114
  },
115
  {
116
  "id": "foundation_model_plan_json",
@@ -120,8 +120,8 @@
120
  "surface": "website_hf",
121
  "shows": "Machine-readable foundation-model selection matrix with source links, entry conditions, and evaluation additions.",
122
  "exists": true,
123
- "bytes": 13193,
124
- "sha256": "63529cbaf1d5c549f595b3ed49f49feda03edf96952b5cb321117fee340849c9"
125
  },
126
  {
127
  "id": "omni_model_extension_contract",
@@ -142,8 +142,8 @@
142
  "surface": "repo_hf",
143
  "shows": "Stores the implemented Qwen3-Omni LoRA contract and planned Cosmos-style world-model and VLA/policy branch contracts.",
144
  "exists": true,
145
- "file_count": 4,
146
- "bytes": 12613
147
  },
148
  {
149
  "id": "omni_backbone_registry_validator",
@@ -211,6 +211,94 @@
211
  "bytes": 48164,
212
  "sha256": "fbefe3f31e2d19566ed1fc356a25c564ecb4f0645de4d595f5926e1426c058d5"
213
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
214
  {
215
  "id": "qwen3_lora_hf_package_builder",
216
  "title": "Qwen3 LoRA HF package builder",
@@ -219,8 +307,8 @@
219
  "surface": "repo_hf",
220
  "shows": "Builds the upload-ready Hugging Face adapter folder from a verified Qwen3 LoRA result summary and adapter directory.",
221
  "exists": true,
222
- "bytes": 10710,
223
- "sha256": "c5f33f7030e6861ccab465b71d4c1b4209b8fbbcebdca882aa5c7e8d70e76b37"
224
  },
225
  {
226
  "id": "additional_development_directions",
@@ -274,8 +362,8 @@
274
  "surface": "website_hf",
275
  "shows": "Gives a short project path with scope status and public surfaces.",
276
  "exists": true,
277
- "bytes": 8005,
278
- "sha256": "2258fecb80850c745e60cb28733869c49a5182879d9d0461b666a5575e3c1610"
279
  },
280
  {
281
  "id": "artifact_guide",
@@ -285,8 +373,8 @@
285
  "surface": "repo_hf",
286
  "shows": "Gives the human-readable map from project scope to data, tasks, platform mirrors, and scale-up status.",
287
  "exists": true,
288
- "bytes": 17930,
289
- "sha256": "6a2843318d83b8dac1fc56a53e4dbf202f8757d4e7d8b3bf67b574be570fa585"
290
  },
291
  {
292
  "id": "official_dataset_card_alignment",
@@ -330,7 +418,7 @@
330
  "shows": "Machine-readable source-alignment pass/fail check for repo, website, and HF surfaces.",
331
  "exists": true,
332
  "bytes": 4432,
333
- "sha256": "06c6e2d111c72df01ed127fd288e6675b63e35a21ae12a2523931a072bd0bc49"
334
  },
335
  {
336
  "id": "source_alignment_validator",
@@ -351,8 +439,8 @@
351
  "surface": "repo_hf",
352
  "shows": "Publishes prepared Space, artifact dataset, and model bundles, including an explicit model-binary upload batch.",
353
  "exists": true,
354
- "bytes": 15927,
355
- "sha256": "7b3e515763ccce08f72b4fd12a903c21f14b469c6af144524196ecad945da2ab"
356
  },
357
  {
358
  "id": "github_package_dockerfile",
@@ -582,8 +670,8 @@
582
  "surface": "website_hf",
583
  "shows": "Machine-readable release-check summary for validators, mirrors, and public project surfaces.",
584
  "exists": true,
585
- "bytes": 8099,
586
- "sha256": "c67b1a3cbd1a355ffdecec22d22486c181f081b03286f4180acff7f9659402e0"
587
  },
588
  {
589
  "id": "public_surface_qa",
@@ -605,7 +693,7 @@
605
  "volatile": true,
606
  "shows": "Machine-readable report for SEO/social metadata, accessible tab semantics, public links, project links, and clear project presentation.",
607
  "exists": true,
608
- "bytes": 5591,
609
  "hash_policy": "existence_and_size_only"
610
  },
611
  {
@@ -616,8 +704,8 @@
616
  "surface": "repo_hf",
617
  "shows": "Regenerates the public presentation report before release.",
618
  "exists": true,
619
- "bytes": 11986,
620
- "sha256": "949fffe589d35db8ba342c44d462702acf00be129f4d3392a239496f797252fa"
621
  },
622
  {
623
  "id": "task_surface_integrity",
@@ -686,7 +774,7 @@
686
  "volatile": true,
687
  "shows": "Records the last live GitHub/HF URL verification after upload.",
688
  "exists": true,
689
- "bytes": 74880,
690
  "hash_policy": "existence_and_size_only"
691
  },
692
  {
@@ -697,8 +785,8 @@
697
  "surface": "repo",
698
  "shows": "Fetches the published GitHub/HF URLs and compares live hashes and public-card markers against the release assets.",
699
  "exists": true,
700
- "bytes": 36201,
701
- "sha256": "76f03885867a8ed7095958a6948cbce81b4958fb74a09df24c24ad7eb5b0d944"
702
  },
703
  {
704
  "id": "reproducibility_contract",
@@ -730,8 +818,8 @@
730
  "surface": "repo_hf",
731
  "shows": "Generates the selective artifact catalog from local files.",
732
  "exists": true,
733
- "bytes": 38561,
734
- "sha256": "571a06684909bd4d544d455d5cdee2fb69439b1e16de95609dd51fecc7b58b29"
735
  },
736
  {
737
  "id": "publication_audit",
@@ -742,7 +830,7 @@
742
  "volatile": true,
743
  "shows": "Confirms public bundles exclude raw data, caches, heavy archives, and credential text.",
744
  "exists": true,
745
- "bytes": 7212,
746
  "hash_policy": "existence_and_size_only"
747
  },
748
  {
@@ -754,7 +842,7 @@
754
  "volatile": true,
755
  "shows": "Separates setup paths from completed held-out-episode results.",
756
  "exists": true,
757
- "bytes": 21251,
758
  "hash_policy": "existence_and_size_only"
759
  },
760
  {
@@ -766,7 +854,7 @@
766
  "volatile": true,
767
  "shows": "Confirms prepared GitHub/HF Space/artifact/model mirrors share the same critical data, figure, website HTML, and validator files.",
768
  "exists": true,
769
- "bytes": 410374,
770
  "hash_policy": "existence_and_size_only"
771
  },
772
  {
@@ -778,7 +866,7 @@
778
  "volatile": true,
779
  "shows": "Confirms local website links, anchors, JSON data files, and referenced images resolve.",
780
  "exists": true,
781
- "bytes": 15375,
782
  "hash_policy": "existence_and_size_only"
783
  },
784
  {
@@ -811,8 +899,8 @@
811
  "surface": "website_hf",
812
  "shows": "Mirrors task metrics for the static dashboard.",
813
  "exists": true,
814
- "bytes": 27490,
815
- "sha256": "159ed565571aa4215ef30a5ea8fce057481cf0f77ad50aec3ae15de6a38e12ba"
816
  },
817
  {
818
  "id": "feature_manifest",
@@ -965,8 +1053,8 @@
965
  "surface": "repo_hf",
966
  "shows": "Documents the final 128-episode LoRA adapter upload path, target model repo, package builder, and forbidden files.",
967
  "exists": true,
968
- "bytes": 1886,
969
- "sha256": "a8fa5a326e2dd91130c2e86b988e7afcfa4e043e48614191f0fc589158089c03"
970
  },
971
  {
972
  "id": "multi_episode_access_status",
@@ -1031,8 +1119,8 @@
1031
  "surface": "repo_hf",
1032
  "shows": "Reader-facing comparison of the single-episode task suite, 128-episode aligned baselines, Qwen3-Omni packages, and Cosmos3 future-window branch.",
1033
  "exists": true,
1034
- "bytes": 9231,
1035
- "sha256": "c38d12e138193f7200800d4dd8c149497de2c5f5895299e22fe81285b69fc62d"
1036
  },
1037
  {
1038
  "id": "omni_model_comparison_json",
@@ -1042,8 +1130,8 @@
1042
  "surface": "repo_hf",
1043
  "shows": "Machine-readable comparison of the current result versions, per-task aligned baselines, verified Qwen3 packages, and Cosmos3 package.",
1044
  "exists": true,
1045
- "bytes": 48296,
1046
- "sha256": "1c968bd58842af9a4e6159c1a8bd171aec08757bb77fce9f04c55030be08357f"
1047
  },
1048
  {
1049
  "id": "cosmos3_nano_verified_summary",
@@ -1144,6 +1232,72 @@
1144
  "bytes": 1099,
1145
  "sha256": "f11ccb167908d4f5bfb49c0be0b4bc6c9254901462aa52ae98a2a98e8af16558"
1146
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1147
  {
1148
  "id": "verified_public_package_xperience10m_cosmos3_super_reasoner_128ep_test_full_20260607",
1149
  "title": "Verified public package: Cosmos3-Super Reasoner",
@@ -1331,6 +1485,61 @@
1331
  "bytes": 669,
1332
  "sha256": "3d427e70e44b22b882be49f2963e2afcf5b497f25c445850c9f567cdbc41ed15"
1333
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1334
  {
1335
  "id": "verified_public_package_xperience10m_qwen3_omni_128ep_structured_json_v2_reuse_full8gpu_lora_eval_test_full",
1336
  "title": "Verified public package: Qwen3-Omni LoRA",
@@ -1462,6 +1671,72 @@
1462
  "exists": true,
1463
  "bytes": 634,
1464
  "sha256": "c5bfa7b231dcf2f9a2d1a50ce5324cdf4bb8afd60294fd5d40b7ef578983b1b2"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1465
  }
1466
  ]
1467
  }
 
1
  {
2
  "title": "Ropedia Xperience-10M Task Suite Artifact Index",
3
+ "generated_at_utc": "2026-06-11T03:34:19+00:00",
4
  "status": "pass",
5
+ "artifact_count": 154,
6
  "missing": [],
7
  "by_kind": {
8
  "project_path": 14,
9
  "scaleup_contract": 7,
10
+ "scaleup_status": 34,
11
  "publication_workflow": 5,
12
  "project_scope": 1,
13
  "source_alignment": 5,
14
  "evaluation_protocol": 3,
15
  "result_interpretation": 5,
16
+ "metrics_source": 24,
17
  "website_data": 3,
18
  "visual_evidence": 7,
19
  "quality_gate": 12,
 
31
  "generated_figure_assets": 1,
32
  "citation": 1,
33
  "license": 1,
34
+ "verified_public_package": 9,
35
+ "publication_audit": 6
36
  },
37
  "artifacts": [
38
  {
 
65
  "surface": "repo_hf",
66
  "shows": "Gives a compact current-state table for first-pass readers.",
67
  "exists": true,
68
+ "bytes": 13535,
69
+ "sha256": "595f90e26d75baba46a0827eb2ec38d0c6badb137437f642dff4e88e22b7ca80"
70
  },
71
  {
72
  "id": "project_status_json",
 
76
  "surface": "website_hf",
77
  "shows": "Machine-readable copy of the current project status for website and HF mirrors.",
78
  "exists": true,
79
+ "bytes": 23198,
80
+ "sha256": "07678354fde90a7e15134f5deff036eb200f7bf422ab2350ff64825b902843d5"
81
  },
82
  {
83
  "id": "research_roadmap",
 
87
  "surface": "repo_hf",
88
  "shows": "Defines the path from public-sample task development to multi-episode held-out evaluation and larger omni-model extensions.",
89
  "exists": true,
90
+ "bytes": 14354,
91
+ "sha256": "7d385916e6d07cfa5f95e80b501fe58237f671711691cc8c792479d233991ef5"
92
  },
93
  {
94
  "id": "research_roadmap_json",
 
98
  "surface": "website_hf",
99
  "shows": "Machine-readable research roadmap for the website and Hugging Face mirrors.",
100
  "exists": true,
101
+ "bytes": 13313,
102
+ "sha256": "719a8b5f45810d6b09788391d8549eb5226e480454c9f2a4ed7380ceacfd834b"
103
  },
104
  {
105
  "id": "foundation_model_plan",
 
109
  "surface": "repo_hf",
110
  "shows": "Defines the post-data-gate backbone choices: Qwen3-Omni first, Cosmos 3 for world modeling, and VLA/policy models after action-target conversion.",
111
  "exists": true,
112
+ "bytes": 9550,
113
+ "sha256": "5adb6a3be92ccde189176cc618eb0814ec63f094191a61f7695d3f55fae36b3d"
114
  },
115
  {
116
  "id": "foundation_model_plan_json",
 
120
  "surface": "website_hf",
121
  "shows": "Machine-readable foundation-model selection matrix with source links, entry conditions, and evaluation additions.",
122
  "exists": true,
123
+ "bytes": 13457,
124
+ "sha256": "7d17f3c6a2cd9954c9632d7063d3aab7d5753c709215239257f065274ab394f3"
125
  },
126
  {
127
  "id": "omni_model_extension_contract",
 
142
  "surface": "repo_hf",
143
  "shows": "Stores the implemented Qwen3-Omni LoRA contract and planned Cosmos-style world-model and VLA/policy branch contracts.",
144
  "exists": true,
145
+ "file_count": 5,
146
+ "bytes": 16418
147
  },
148
  {
149
  "id": "omni_backbone_registry_validator",
 
211
  "bytes": 48164,
212
  "sha256": "fbefe3f31e2d19566ed1fc356a25c564ecb4f0645de4d595f5926e1426c058d5"
213
  },
214
+ {
215
+ "id": "task_suite_enhancement_128",
216
+ "title": "128-episode task-suite enhancement pack",
217
+ "path": "TASK_SUITE_ENHANCEMENT_128.md",
218
+ "kind": "scaleup_status",
219
+ "surface": "repo_hf",
220
+ "shows": "Records the no-new-episode dense-window, hierarchical-target, bottleneck, and experiment-backlog plan for pushing the current 128-episode suite harder without overwriting prior results.",
221
+ "exists": true,
222
+ "bytes": 3427,
223
+ "sha256": "f6b30c2c288ed2dd2b1b1bb36c334d9697665201c60f77374d38b1ff3911fc86"
224
+ },
225
+ {
226
+ "id": "task_suite_enhancement_128_json",
227
+ "title": "128-episode task-suite enhancement JSON",
228
+ "path": "docs/data/task_suite_enhancement_128.json",
229
+ "kind": "scaleup_status",
230
+ "surface": "website_hf",
231
+ "shows": "Machine-readable enhancement pack for the website and Hugging Face mirrors.",
232
+ "exists": true,
233
+ "bytes": 20181,
234
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e"
235
+ },
236
+ {
237
+ "id": "task_suite_enhancement_128_result",
238
+ "title": "128-episode task-suite enhancement result package",
239
+ "path": "results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json",
240
+ "kind": "scaleup_status",
241
+ "surface": "repo_hf",
242
+ "shows": "Versioned result directory with dense-window estimates, hierarchical target contract, task bottlenecks, Qwen action-family error summary, and experiment cards.",
243
+ "exists": true,
244
+ "bytes": 20181,
245
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e"
246
+ },
247
+ {
248
+ "id": "task_suite_enhancement_128_builder",
249
+ "title": "128-episode task-suite enhancement builder",
250
+ "path": "scripts/omni/build_task_suite_enhancement_128.py",
251
+ "kind": "scaleup_status",
252
+ "surface": "repo_hf",
253
+ "shows": "Regenerates the enhancement pack from committed 128-episode windows, baseline summaries, verified Qwen predictions, and Cosmos reference metrics.",
254
+ "exists": true,
255
+ "bytes": 27210,
256
+ "sha256": "0e098d7c1a5c91ec8472d5eb8fc0ebab0305cf647d1f0f4f2ba6bd4c1d531546"
257
+ },
258
+ {
259
+ "id": "qwen3_full_parameter_gates",
260
+ "title": "Qwen3-Omni full-parameter feasibility gates",
261
+ "path": "results/omni_finetune/QWEN3_FULL_PARAMETER_GATES_20260609.md",
262
+ "kind": "scaleup_status",
263
+ "surface": "repo_hf",
264
+ "shows": "Summarizes the 2026-06-09 full-parameter FSDP feasibility gates: 1/8/32/64-step guarded runs passed, the 128-step opportunistic pilot was preempted for Qwen v5 handoff, and no full checkpoints or weights are published.",
265
+ "exists": true,
266
+ "bytes": 3003,
267
+ "sha256": "5c363f9f5e6be2bcf5bc7b65f7ad7590c86faf31392a9a5bd74987dc5931d7eb"
268
+ },
269
+ {
270
+ "id": "qwen3_full_parameter_gates_json",
271
+ "title": "Qwen3-Omni full-parameter feasibility gates JSON",
272
+ "path": "docs/data/qwen3_full_parameter_gates.json",
273
+ "kind": "scaleup_status",
274
+ "surface": "website_hf",
275
+ "shows": "Machine-readable summary of full-parameter feasibility evidence and publication policy for website and Hugging Face mirrors.",
276
+ "exists": true,
277
+ "bytes": 10533,
278
+ "sha256": "7eecba00a26aa9df04048816a112011b5d0182b3fbd1aee19686cf3eb7828fd5"
279
+ },
280
+ {
281
+ "id": "qwen3_full_parameter_gates_builder",
282
+ "title": "Qwen3-Omni full-parameter gate summary builder",
283
+ "path": "scripts/omni/build_qwen3_full_parameter_gate_summary.py",
284
+ "kind": "scaleup_status",
285
+ "surface": "repo_hf",
286
+ "shows": "Regenerates the full-parameter feasibility-gate Markdown and JSON summaries from the run-local evidence files.",
287
+ "exists": true,
288
+ "bytes": 15325,
289
+ "sha256": "55ad8af43484d92d98bd3200ea781929397752cdcc2bfbd8e8160a46cb8a5730"
290
+ },
291
+ {
292
+ "id": "qwen3_full_parameter_post_verified_deferrer",
293
+ "title": "Qwen3-Omni post-verified full-parameter deferrer",
294
+ "path": "scripts/omni/defer_qwen3_fullparam_after_verified_qwen.sh",
295
+ "kind": "scaleup_status",
296
+ "surface": "repo_hf",
297
+ "shows": "Waits for a verified Qwen held-out package, then launches a bounded 128-step full-parameter feasibility pilot on the same multiscale v5 dataset with no checkpoints or weights saved.",
298
+ "exists": true,
299
+ "bytes": 1840,
300
+ "sha256": "db088d179b8f35abea6cd10ed71f52da5ecf00b3c7ed9c2e65880450553d8e06"
301
+ },
302
  {
303
  "id": "qwen3_lora_hf_package_builder",
304
  "title": "Qwen3 LoRA HF package builder",
 
307
  "surface": "repo_hf",
308
  "shows": "Builds the upload-ready Hugging Face adapter folder from a verified Qwen3 LoRA result summary and adapter directory.",
309
  "exists": true,
310
+ "bytes": 11332,
311
+ "sha256": "e9f9909185c5c647a5df2444d0ce2c3d07dca719fdd92c9bf1c7964e19963fa1"
312
  },
313
  {
314
  "id": "additional_development_directions",
 
362
  "surface": "website_hf",
363
  "shows": "Gives a short project path with scope status and public surfaces.",
364
  "exists": true,
365
+ "bytes": 10496,
366
+ "sha256": "fc21325035c3b8d4892bfad50dd1a7f5f562a662b7da7be73f7349cb515544ef"
367
  },
368
  {
369
  "id": "artifact_guide",
 
373
  "surface": "repo_hf",
374
  "shows": "Gives the human-readable map from project scope to data, tasks, platform mirrors, and scale-up status.",
375
  "exists": true,
376
+ "bytes": 19019,
377
+ "sha256": "b50c681fc3b9e1255d38631c0ebe9d4653a8acb0607b3455f9f827638b9f65e6"
378
  },
379
  {
380
  "id": "official_dataset_card_alignment",
 
418
  "shows": "Machine-readable source-alignment pass/fail check for repo, website, and HF surfaces.",
419
  "exists": true,
420
  "bytes": 4432,
421
+ "sha256": "e3644aa1fa819807688607e56e0bf1d98a84c9a187f58a6b234a73ea8b25bd0f"
422
  },
423
  {
424
  "id": "source_alignment_validator",
 
439
  "surface": "repo_hf",
440
  "shows": "Publishes prepared Space, artifact dataset, and model bundles, including an explicit model-binary upload batch.",
441
  "exists": true,
442
+ "bytes": 19352,
443
+ "sha256": "b7eb9bf2abb842bd2f411b395ba6f75bb0a98b86fa35bdedc68430855f0f7da9"
444
  },
445
  {
446
  "id": "github_package_dockerfile",
 
670
  "surface": "website_hf",
671
  "shows": "Machine-readable release-check summary for validators, mirrors, and public project surfaces.",
672
  "exists": true,
673
+ "bytes": 8097,
674
+ "sha256": "69ffacd5860853a4b46717b2ea7b637e10ef782be8e809d91371707d7b29c90d"
675
  },
676
  {
677
  "id": "public_surface_qa",
 
693
  "volatile": true,
694
  "shows": "Machine-readable report for SEO/social metadata, accessible tab semantics, public links, project links, and clear project presentation.",
695
  "exists": true,
696
+ "bytes": 5640,
697
  "hash_policy": "existence_and_size_only"
698
  },
699
  {
 
704
  "surface": "repo_hf",
705
  "shows": "Regenerates the public presentation report before release.",
706
  "exists": true,
707
+ "bytes": 12034,
708
+ "sha256": "1a2f23dc2be5dfa3a10438162579b1304f7f069e333a5a6c715f9001b95ac2a9"
709
  },
710
  {
711
  "id": "task_surface_integrity",
 
774
  "volatile": true,
775
  "shows": "Records the last live GitHub/HF URL verification after upload.",
776
  "exists": true,
777
+ "bytes": 89685,
778
  "hash_policy": "existence_and_size_only"
779
  },
780
  {
 
785
  "surface": "repo",
786
  "shows": "Fetches the published GitHub/HF URLs and compares live hashes and public-card markers against the release assets.",
787
  "exists": true,
788
+ "bytes": 42224,
789
+ "sha256": "069b05229b72b9fa0c0917afd18655f58e5258499b43889dd96a87f29f577890"
790
  },
791
  {
792
  "id": "reproducibility_contract",
 
818
  "surface": "repo_hf",
819
  "shows": "Generates the selective artifact catalog from local files.",
820
  "exists": true,
821
+ "bytes": 42045,
822
+ "sha256": "0954119cda59bb9c77cba362a206ac237eae73c4a1dc44cfa56599f170fc94de"
823
  },
824
  {
825
  "id": "publication_audit",
 
830
  "volatile": true,
831
  "shows": "Confirms public bundles exclude raw data, caches, heavy archives, and credential text.",
832
  "exists": true,
833
+ "bytes": 7566,
834
  "hash_policy": "existence_and_size_only"
835
  },
836
  {
 
842
  "volatile": true,
843
  "shows": "Separates setup paths from completed held-out-episode results.",
844
  "exists": true,
845
+ "bytes": 21325,
846
  "hash_policy": "existence_and_size_only"
847
  },
848
  {
 
854
  "volatile": true,
855
  "shows": "Confirms prepared GitHub/HF Space/artifact/model mirrors share the same critical data, figure, website HTML, and validator files.",
856
  "exists": true,
857
+ "bytes": 495491,
858
  "hash_policy": "existence_and_size_only"
859
  },
860
  {
 
866
  "volatile": true,
867
  "shows": "Confirms local website links, anchors, JSON data files, and referenced images resolve.",
868
  "exists": true,
869
+ "bytes": 15654,
870
  "hash_policy": "existence_and_size_only"
871
  },
872
  {
 
899
  "surface": "website_hf",
900
  "shows": "Mirrors task metrics for the static dashboard.",
901
  "exists": true,
902
+ "bytes": 27604,
903
+ "sha256": "ebaf9d598b4cd91118f149cafa01fe4d17629499565b7be0c7ce4a0ffcd70f6b"
904
  },
905
  {
906
  "id": "feature_manifest",
 
1053
  "surface": "repo_hf",
1054
  "shows": "Documents the final 128-episode LoRA adapter upload path, target model repo, package builder, and forbidden files.",
1055
  "exists": true,
1056
+ "bytes": 1877,
1057
+ "sha256": "d12d0b15eb6374b380940a95d13d87c8f840bb5118cf4b57e57965e018ec14e6"
1058
  },
1059
  {
1060
  "id": "multi_episode_access_status",
 
1119
  "surface": "repo_hf",
1120
  "shows": "Reader-facing comparison of the single-episode task suite, 128-episode aligned baselines, Qwen3-Omni packages, and Cosmos3 future-window branch.",
1121
  "exists": true,
1122
+ "bytes": 14840,
1123
+ "sha256": "a3831ff27ed8aecda6e117c228fc1303ab819458eed835fb3b73c7babeb8e136"
1124
  },
1125
  {
1126
  "id": "omni_model_comparison_json",
 
1130
  "surface": "repo_hf",
1131
  "shows": "Machine-readable comparison of the current result versions, per-task aligned baselines, verified Qwen3 packages, and Cosmos3 package.",
1132
  "exists": true,
1133
+ "bytes": 75698,
1134
+ "sha256": "3b310ccf4e1c72aba1382f051778d581fd150cef468db826869ade8e8149abca"
1135
  },
1136
  {
1137
  "id": "cosmos3_nano_verified_summary",
 
1232
  "bytes": 1099,
1233
  "sha256": "f11ccb167908d4f5bfb49c0be0b4bc6c9254901462aa52ae98a2a98e8af16558"
1234
  },
1235
+ {
1236
+ "id": "verified_public_package_xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1237
+ "title": "Verified public package: Cosmos3-Super Forward-Dynamics LoRA",
1238
+ "path": "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1239
+ "kind": "verified_public_package",
1240
+ "surface": "repo_hf",
1241
+ "shows": "Public-safe verified package for xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp (cosmos3_super_forward_dynamics, status=verified).",
1242
+ "exists": true,
1243
+ "file_count": 31,
1244
+ "bytes": 1868234
1245
+ },
1246
+ {
1247
+ "id": "verified_public_summary_xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1248
+ "title": "Verified summary: Cosmos3-Super Forward-Dynamics LoRA",
1249
+ "path": "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/verified_result_summary.json",
1250
+ "kind": "metrics_source",
1251
+ "surface": "repo_hf",
1252
+ "shows": "Machine-readable verified summary for xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp.",
1253
+ "exists": true,
1254
+ "bytes": 3978,
1255
+ "sha256": "f70d7507537abcf9ac140f0f1b44d0dad7c5b324cc35635fd70c8f6f85d4ae94"
1256
+ },
1257
+ {
1258
+ "id": "verified_public_public_result_summary_xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1259
+ "title": "Verified public result summary: Cosmos3-Super Forward-Dynamics LoRA",
1260
+ "path": "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/PUBLIC_RESULT_SUMMARY.md",
1261
+ "kind": "scaleup_status",
1262
+ "surface": "repo_hf",
1263
+ "shows": "Public result summary for xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp.",
1264
+ "exists": true,
1265
+ "bytes": 860,
1266
+ "sha256": "15ab90a41a708629865feff9380b0b6c55a68adf178384c4840ed88d3fa1eeeb"
1267
+ },
1268
+ {
1269
+ "id": "verified_public_run_report_xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1270
+ "title": "Verified run report: Cosmos3-Super Forward-Dynamics LoRA",
1271
+ "path": "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/eval/RUN_REPORT.md",
1272
+ "kind": "scaleup_status",
1273
+ "surface": "repo_hf",
1274
+ "shows": "Run report for xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp.",
1275
+ "exists": true,
1276
+ "bytes": 598,
1277
+ "sha256": "ab1874953b1fe4fdfbe38181e4181114797f25b5e8b983cdf74cf6b7475ab111"
1278
+ },
1279
+ {
1280
+ "id": "verified_public_metrics_JSON_xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1281
+ "title": "Verified metrics JSON: Cosmos3-Super Forward-Dynamics LoRA",
1282
+ "path": "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/eval/metrics.json",
1283
+ "kind": "metrics_source",
1284
+ "surface": "repo_hf",
1285
+ "shows": "Metrics json for xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp.",
1286
+ "exists": true,
1287
+ "bytes": 142983,
1288
+ "sha256": "40c0863faead8e0a852fd1efb3ba6f9e97cc8dd8327dd0d3a9fd12f9fff3b615"
1289
+ },
1290
+ {
1291
+ "id": "verified_public_package_audit_xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1292
+ "title": "Verified package audit: Cosmos3-Super Forward-Dynamics LoRA",
1293
+ "path": "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/package_audit.json",
1294
+ "kind": "publication_audit",
1295
+ "surface": "repo_hf",
1296
+ "shows": "Package audit for xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp.",
1297
+ "exists": true,
1298
+ "bytes": 534,
1299
+ "sha256": "9d63e5f0febe52f1dd78a67b3ef4db10b3ecc40c2ddd704b73b03995b220e35e"
1300
+ },
1301
  {
1302
  "id": "verified_public_package_xperience10m_cosmos3_super_reasoner_128ep_test_full_20260607",
1303
  "title": "Verified public package: Cosmos3-Super Reasoner",
 
1485
  "bytes": 669,
1486
  "sha256": "3d427e70e44b22b882be49f2963e2afcf5b497f25c445850c9f567cdbc41ed15"
1487
  },
1488
+ {
1489
+ "id": "verified_public_package_xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
1490
+ "title": "Verified public package: Qwen3-Omni LoRA",
1491
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
1492
+ "kind": "verified_public_package",
1493
+ "surface": "repo_hf",
1494
+ "shows": "Public-safe verified package for xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full (qwen3_omni_lora, status=verified).",
1495
+ "exists": true,
1496
+ "file_count": 13,
1497
+ "bytes": 11998803
1498
+ },
1499
+ {
1500
+ "id": "verified_public_summary_xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
1501
+ "title": "Verified summary: Qwen3-Omni LoRA",
1502
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full/verified_result_summary.json",
1503
+ "kind": "metrics_source",
1504
+ "surface": "repo_hf",
1505
+ "shows": "Machine-readable verified summary for xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full.",
1506
+ "exists": true,
1507
+ "bytes": 5909,
1508
+ "sha256": "2f05264c8d3974545fc48074ba6c2a60741fe478aef6298ba3c3e35f3998967a"
1509
+ },
1510
+ {
1511
+ "id": "verified_public_public_result_summary_xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
1512
+ "title": "Verified public result summary: Qwen3-Omni LoRA",
1513
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full/PUBLIC_RESULT_SUMMARY.md",
1514
+ "kind": "scaleup_status",
1515
+ "surface": "repo_hf",
1516
+ "shows": "Public result summary for xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full.",
1517
+ "exists": true,
1518
+ "bytes": 991,
1519
+ "sha256": "51ac77b38e7b3231bd9be70c78a8f3f3632b3022360b8651dea5e74a8378a487"
1520
+ },
1521
+ {
1522
+ "id": "verified_public_run_report_xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
1523
+ "title": "Verified run report: Qwen3-Omni LoRA",
1524
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full/eval/RUN_REPORT.md",
1525
+ "kind": "scaleup_status",
1526
+ "surface": "repo_hf",
1527
+ "shows": "Run report for xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full.",
1528
+ "exists": true,
1529
+ "bytes": 593,
1530
+ "sha256": "29d3839f384815c7736eb1c14c697c2ef1984c8a58c03d4354510fc60a4185ae"
1531
+ },
1532
+ {
1533
+ "id": "verified_public_metrics_JSON_xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
1534
+ "title": "Verified metrics JSON: Qwen3-Omni LoRA",
1535
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full/eval/metrics.json",
1536
+ "kind": "metrics_source",
1537
+ "surface": "repo_hf",
1538
+ "shows": "Metrics json for xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full.",
1539
+ "exists": true,
1540
+ "bytes": 48231,
1541
+ "sha256": "18000afe3d23bd5aad93743f6f86fc3c60c344a69f0cecee9b552e5fd14b8eb5"
1542
+ },
1543
  {
1544
  "id": "verified_public_package_xperience10m_qwen3_omni_128ep_structured_json_v2_reuse_full8gpu_lora_eval_test_full",
1545
  "title": "Verified public package: Qwen3-Omni LoRA",
 
1671
  "exists": true,
1672
  "bytes": 634,
1673
  "sha256": "c5bfa7b231dcf2f9a2d1a50ce5324cdf4bb8afd60294fd5d40b7ef578983b1b2"
1674
+ },
1675
+ {
1676
+ "id": "verified_public_package_xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1677
+ "title": "Verified public package: Qwen3-Omni LoRA",
1678
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1679
+ "kind": "verified_public_package",
1680
+ "surface": "repo_hf",
1681
+ "shows": "Public-safe verified package for xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full (qwen3_omni_lora, status=verified).",
1682
+ "exists": true,
1683
+ "file_count": 16,
1684
+ "bytes": 4854319
1685
+ },
1686
+ {
1687
+ "id": "verified_public_summary_xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1688
+ "title": "Verified summary: Qwen3-Omni LoRA",
1689
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/verified_result_summary.json",
1690
+ "kind": "metrics_source",
1691
+ "surface": "repo_hf",
1692
+ "shows": "Machine-readable verified summary for xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full.",
1693
+ "exists": true,
1694
+ "bytes": 6482,
1695
+ "sha256": "91a70ed837584bba83ca99affed643cb9e50a12ad779f127554e52a436f712e5"
1696
+ },
1697
+ {
1698
+ "id": "verified_public_public_result_summary_xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1699
+ "title": "Verified public result summary: Qwen3-Omni LoRA",
1700
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/PUBLIC_RESULT_SUMMARY.md",
1701
+ "kind": "scaleup_status",
1702
+ "surface": "repo_hf",
1703
+ "shows": "Public result summary for xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full.",
1704
+ "exists": true,
1705
+ "bytes": 989,
1706
+ "sha256": "4b9467e7bf22fa8ff2b3fbd63693df6c38cb752ff881876060cf27dbc911e06c"
1707
+ },
1708
+ {
1709
+ "id": "verified_public_run_report_xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1710
+ "title": "Verified run report: Qwen3-Omni LoRA",
1711
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/eval/RUN_REPORT.md",
1712
+ "kind": "scaleup_status",
1713
+ "surface": "repo_hf",
1714
+ "shows": "Run report for xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full.",
1715
+ "exists": true,
1716
+ "bytes": 443,
1717
+ "sha256": "4143378be332bf125dbbfc65aac5d28ece185cda760a8175f114de047b39e424"
1718
+ },
1719
+ {
1720
+ "id": "verified_public_metrics_JSON_xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1721
+ "title": "Verified metrics JSON: Qwen3-Omni LoRA",
1722
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/eval/metrics.json",
1723
+ "kind": "metrics_source",
1724
+ "surface": "repo_hf",
1725
+ "shows": "Metrics json for xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full.",
1726
+ "exists": true,
1727
+ "bytes": 47584,
1728
+ "sha256": "a1c1f81d4bb53a49b4e912450159c9872de66784aab4f3d41e20c40055daa11e"
1729
+ },
1730
+ {
1731
+ "id": "verified_public_package_audit_xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1732
+ "title": "Verified package audit: Qwen3-Omni LoRA",
1733
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/package_audit.json",
1734
+ "kind": "publication_audit",
1735
+ "surface": "repo_hf",
1736
+ "shows": "Package audit for xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full.",
1737
+ "exists": true,
1738
+ "bytes": 624,
1739
+ "sha256": "0723ecb9dd5cef371e2120af10903fc84a3ef495edafc8959f5eec0a778abbf1"
1740
  }
1741
  ]
1742
  }
data/foundation_model_plan.json CHANGED
@@ -1,14 +1,19 @@
1
  {
2
  "title": "Xperience-10M Foundation Model Plan",
3
  "status": "planning_artifact",
4
- "current_boundary": "A final held-out multi-episode Qwen3-Omni diagnostic result is verified in this repo and meets the strict-JSON target, but it is not a strong action/subtask model result. The current foundation-model work should treat it as the baseline train/eval/package loop before Qwen action/subtask improvements, Cosmos-style world modeling, or policy/VLA branches.",
5
  "backbone_registry": {
6
  "config_dir": "configs/omni_backbones",
7
  "validator": "scripts/omni/backbone_registry.py --validate --json",
8
  "extension_contract": "OMNI_MODEL_EXTENSION_CONTRACT.md",
9
  "implemented_backbone": "qwen3_omni_lora",
10
- "planned_backbones": [
 
11
  "cosmos_world_model",
 
 
 
 
12
  "policy_vla_branch"
13
  ]
14
  },
@@ -107,8 +112,8 @@
107
  "Uses pose/SLAM, depth, mocap, IMU, and language as physical-world conditioning signals.",
108
  "Better aligned with prediction/generation objectives than simple label classification."
109
  ],
110
- "current_decision": "add_as_first_world_model_branch_after_data_gate",
111
- "entry_condition": "Multi-episode data plus enough storage/compute for generated or latent video-state outputs.",
112
  "public_source": "https://www.nvidia.com/en-us/ai/cosmos/"
113
  },
114
  {
@@ -211,12 +216,12 @@
211
  {
212
  "step": 3,
213
  "name": "Model-selection dry run",
214
- "action": "Run 3-8 episode dry runs for Qwen3-Omni prompt/LoRA, Cosmos 3 preprocessing, and one policy candidate."
215
  },
216
  {
217
  "step": 4,
218
  "name": "World-model branch",
219
- "action": "Promote Cosmos 3 if future-window/action-conditioned preprocessing fits storage and compute."
220
  },
221
  {
222
  "step": 5,
 
1
  {
2
  "title": "Xperience-10M Foundation Model Plan",
3
  "status": "planning_artifact",
4
+ "current_boundary": "Verified held-out multi-episode foundation-model diagnostics now exist for Qwen3-Omni LoRA, Cosmos3-Nano future-window compatibility, Cosmos3-Super base-weight Reasoner evaluation, and Cosmos3-Super Forward-Dynamics LoRA. Qwen remains the structured JSON baseline; the Cosmos branches answer world-model or base-reasoner questions with separate metrics.",
5
  "backbone_registry": {
6
  "config_dir": "configs/omni_backbones",
7
  "validator": "scripts/omni/backbone_registry.py --validate --json",
8
  "extension_contract": "OMNI_MODEL_EXTENSION_CONTRACT.md",
9
  "implemented_backbone": "qwen3_omni_lora",
10
+ "implemented_backbones": [
11
+ "qwen3_omni_lora",
12
  "cosmos_world_model",
13
+ "cosmos3_super_reasoner",
14
+ "cosmos3_super_forward_dynamics"
15
+ ],
16
+ "planned_backbones": [
17
  "policy_vla_branch"
18
  ]
19
  },
 
112
  "Uses pose/SLAM, depth, mocap, IMU, and language as physical-world conditioning signals.",
113
  "Better aligned with prediction/generation objectives than simple label classification."
114
  ],
115
+ "current_decision": "implemented_as_nano_future_window_and_super_forward_dynamics_branches",
116
+ "entry_condition": "Use separate metrics for Nano future-window retrieval and Super forward-dynamics MSE; do not compare them directly to Qwen JSON-task accuracy.",
117
  "public_source": "https://www.nvidia.com/en-us/ai/cosmos/"
118
  },
119
  {
 
216
  {
217
  "step": 3,
218
  "name": "Model-selection dry run",
219
+ "action": "Run 3-8 episode dry runs for any next backbone before scaling beyond the selected split."
220
  },
221
  {
222
  "step": 4,
223
  "name": "World-model branch",
224
+ "action": "Promote Cosmos 3 beyond the current Nano compatibility and Super forward-dynamics runs only when loss metrics, preprocessing, and storage justify the added compute."
225
  },
226
  {
227
  "step": 5,
data/live_publication_status.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "title": "Ropedia Xperience-10M Live Publication Status",
3
  "status": "pass",
4
- "checked_at_utc": "2026-06-07T09:49:27+00:00",
5
- "scope": "Live GitHub Pages, GitHub raw, Hugging Face Space, artifact dataset, baseline model mirrors, and the Qwen3 LoRA adapter repo when the final upload package exists locally.",
6
  "hash_groups": [
7
  {
8
  "id": "task_suite_infographic",
@@ -17,6 +17,8 @@
17
  "mirrors": {
18
  "github_pages": {
19
  "ok": true,
 
 
20
  "status_code": 200,
21
  "bytes": 1588641,
22
  "sha256": "1275e2adaef920ecde7c29dc62c8d79d4f13475a0c09bc3baa693f47cdec2e1f",
@@ -25,6 +27,8 @@
25
  },
26
  "hf_space": {
27
  "ok": true,
 
 
28
  "status_code": 200,
29
  "bytes": 1588641,
30
  "sha256": "1275e2adaef920ecde7c29dc62c8d79d4f13475a0c09bc3baa693f47cdec2e1f",
@@ -33,6 +37,8 @@
33
  },
34
  "hf_artifacts": {
35
  "ok": true,
 
 
36
  "status_code": 200,
37
  "bytes": 1588641,
38
  "sha256": "1275e2adaef920ecde7c29dc62c8d79d4f13475a0c09bc3baa693f47cdec2e1f",
@@ -41,6 +47,8 @@
41
  },
42
  "hf_model": {
43
  "ok": true,
 
 
44
  "status_code": 200,
45
  "bytes": 1588641,
46
  "sha256": "1275e2adaef920ecde7c29dc62c8d79d4f13475a0c09bc3baa693f47cdec2e1f",
@@ -57,40 +65,48 @@
57
  "local": {
58
  "path": "docs/data/quality_gates.json",
59
  "exists": true,
60
- "bytes": 8099,
61
- "sha256": "c67b1a3cbd1a355ffdecec22d22486c181f081b03286f4180acff7f9659402e0"
62
  },
63
  "mirrors": {
64
  "github_pages": {
65
  "ok": true,
 
 
66
  "status_code": 200,
67
- "bytes": 8099,
68
- "sha256": "c67b1a3cbd1a355ffdecec22d22486c181f081b03286f4180acff7f9659402e0",
69
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/quality_gates.json",
70
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/quality_gates.json"
71
  },
72
  "hf_space": {
73
  "ok": true,
 
 
74
  "status_code": 200,
75
- "bytes": 8099,
76
- "sha256": "c67b1a3cbd1a355ffdecec22d22486c181f081b03286f4180acff7f9659402e0",
77
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/quality_gates.json",
78
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/quality_gates.json"
79
  },
80
  "hf_artifacts": {
81
  "ok": true,
 
 
82
  "status_code": 200,
83
- "bytes": 8099,
84
- "sha256": "c67b1a3cbd1a355ffdecec22d22486c181f081b03286f4180acff7f9659402e0",
85
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fquality_gates.json",
86
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/quality_gates.json"
87
  },
88
  "hf_model": {
89
  "ok": true,
 
 
90
  "status_code": 200,
91
- "bytes": 8099,
92
- "sha256": "c67b1a3cbd1a355ffdecec22d22486c181f081b03286f4180acff7f9659402e0",
93
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fquality_gates.json",
94
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/quality_gates.json"
95
  }
96
  },
@@ -109,6 +125,8 @@
109
  "mirrors": {
110
  "github_pages": {
111
  "ok": true,
 
 
112
  "status_code": 200,
113
  "bytes": 4305527,
114
  "sha256": "286230d8070150e09595797569a447308a5110241b69edf52e535e3d8a34f827",
@@ -117,6 +135,8 @@
117
  },
118
  "hf_space": {
119
  "ok": true,
 
 
120
  "status_code": 200,
121
  "bytes": 4305527,
122
  "sha256": "286230d8070150e09595797569a447308a5110241b69edf52e535e3d8a34f827",
@@ -125,18 +145,22 @@
125
  },
126
  "hf_artifacts": {
127
  "ok": true,
 
 
128
  "status_code": 200,
129
  "bytes": 4305527,
130
  "sha256": "286230d8070150e09595797569a447308a5110241b69edf52e535e3d8a34f827",
131
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fsingle_episode_explorer.json",
132
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/single_episode_explorer.json"
133
  },
134
  "hf_model": {
135
  "ok": true,
 
 
136
  "status_code": 200,
137
  "bytes": 4305527,
138
  "sha256": "286230d8070150e09595797569a447308a5110241b69edf52e535e3d8a34f827",
139
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fsingle_episode_explorer.json",
140
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/single_episode_explorer.json"
141
  }
142
  },
@@ -155,6 +179,8 @@
155
  "mirrors": {
156
  "github_pages": {
157
  "ok": true,
 
 
158
  "status_code": 200,
159
  "bytes": 2770710,
160
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
@@ -163,6 +189,8 @@
163
  },
164
  "hf_space": {
165
  "ok": true,
 
 
166
  "status_code": 200,
167
  "bytes": 2770710,
168
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
@@ -171,18 +199,22 @@
171
  },
172
  "hf_artifacts": {
173
  "ok": true,
 
 
174
  "status_code": 200,
175
  "bytes": 2770710,
176
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
177
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fsingle_episode_explorer.html",
178
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/single_episode_explorer.html"
179
  },
180
  "hf_model": {
181
  "ok": true,
 
 
182
  "status_code": 200,
183
  "bytes": 2770710,
184
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
185
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/single_episode_explorer.html",
186
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/single_episode_explorer.html"
187
  }
188
  },
@@ -201,6 +233,8 @@
201
  "mirrors": {
202
  "github_pages": {
203
  "ok": true,
 
 
204
  "status_code": 200,
205
  "bytes": 31706,
206
  "sha256": "a51ccc9e42247be6cda5a026c96be74e2aeeacfae6ef47abf028c9f769eae9a1",
@@ -209,6 +243,8 @@
209
  },
210
  "hf_space": {
211
  "ok": true,
 
 
212
  "status_code": 200,
213
  "bytes": 31706,
214
  "sha256": "a51ccc9e42247be6cda5a026c96be74e2aeeacfae6ef47abf028c9f769eae9a1",
@@ -217,18 +253,22 @@
217
  },
218
  "hf_artifacts": {
219
  "ok": true,
 
 
220
  "status_code": 200,
221
  "bytes": 31706,
222
  "sha256": "a51ccc9e42247be6cda5a026c96be74e2aeeacfae6ef47abf028c9f769eae9a1",
223
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fresearch_roadmap.html",
224
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/research_roadmap.html"
225
  },
226
  "hf_model": {
227
  "ok": true,
 
 
228
  "status_code": 200,
229
  "bytes": 31706,
230
  "sha256": "a51ccc9e42247be6cda5a026c96be74e2aeeacfae6ef47abf028c9f769eae9a1",
231
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/research_roadmap.html",
232
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/research_roadmap.html"
233
  }
234
  },
@@ -247,6 +287,8 @@
247
  "mirrors": {
248
  "github_raw": {
249
  "ok": true,
 
 
250
  "status_code": 200,
251
  "bytes": 4072,
252
  "sha256": "7189be7fc773943859dc75a6119c1fb654d8905ea381e8fc9f782d4b33f5e466",
@@ -255,6 +297,8 @@
255
  },
256
  "hf_space": {
257
  "ok": true,
 
 
258
  "status_code": 200,
259
  "bytes": 4072,
260
  "sha256": "7189be7fc773943859dc75a6119c1fb654d8905ea381e8fc9f782d4b33f5e466",
@@ -263,18 +307,22 @@
263
  },
264
  "hf_artifacts": {
265
  "ok": true,
 
 
266
  "status_code": 200,
267
  "bytes": 4072,
268
  "sha256": "7189be7fc773943859dc75a6119c1fb654d8905ea381e8fc9f782d4b33f5e466",
269
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/results%2Fsingle_episode_diagnostics%2Fprovenance.json",
270
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/results/single_episode_diagnostics/provenance.json"
271
  },
272
  "hf_model": {
273
  "ok": true,
 
 
274
  "status_code": 200,
275
  "bytes": 4072,
276
  "sha256": "7189be7fc773943859dc75a6119c1fb654d8905ea381e8fc9f782d4b33f5e466",
277
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/results%2Fsingle_episode_diagnostics%2Fprovenance.json",
278
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/results/single_episode_diagnostics/provenance.json"
279
  }
280
  },
@@ -293,6 +341,8 @@
293
  "mirrors": {
294
  "github_raw": {
295
  "ok": true,
 
 
296
  "status_code": 200,
297
  "bytes": 995,
298
  "sha256": "813320bbe5d57c74803f5dfb080d440ef1804af41950e031d82ab80311eb04c7",
@@ -301,6 +351,8 @@
301
  },
302
  "hf_space": {
303
  "ok": true,
 
 
304
  "status_code": 200,
305
  "bytes": 995,
306
  "sha256": "813320bbe5d57c74803f5dfb080d440ef1804af41950e031d82ab80311eb04c7",
@@ -309,18 +361,22 @@
309
  },
310
  "hf_artifacts": {
311
  "ok": true,
 
 
312
  "status_code": 200,
313
  "bytes": 995,
314
  "sha256": "813320bbe5d57c74803f5dfb080d440ef1804af41950e031d82ab80311eb04c7",
315
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/results%2Fsingle_episode_diagnostics%2Fobject_labels%2Fobject_vocab.json",
316
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/results/single_episode_diagnostics/object_labels/object_vocab.json"
317
  },
318
  "hf_model": {
319
  "ok": true,
 
 
320
  "status_code": 200,
321
  "bytes": 995,
322
  "sha256": "813320bbe5d57c74803f5dfb080d440ef1804af41950e031d82ab80311eb04c7",
323
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/results%2Fsingle_episode_diagnostics%2Fobject_labels%2Fobject_vocab.json",
324
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/results/single_episode_diagnostics/object_labels/object_vocab.json"
325
  }
326
  },
@@ -333,40 +389,48 @@
333
  "local": {
334
  "path": "docs/data/public_surface_qa.json",
335
  "exists": true,
336
- "bytes": 5591,
337
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1"
338
  },
339
  "mirrors": {
340
  "github_pages": {
341
  "ok": true,
 
 
342
  "status_code": 200,
343
- "bytes": 5591,
344
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1",
345
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/public_surface_qa.json",
346
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/public_surface_qa.json"
347
  },
348
  "hf_space": {
349
  "ok": true,
 
 
350
  "status_code": 200,
351
- "bytes": 5591,
352
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1",
353
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/public_surface_qa.json",
354
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/public_surface_qa.json"
355
  },
356
  "hf_artifacts": {
357
  "ok": true,
 
 
358
  "status_code": 200,
359
- "bytes": 5591,
360
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1",
361
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fpublic_surface_qa.json",
362
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/public_surface_qa.json"
363
  },
364
  "hf_model": {
365
  "ok": true,
 
 
366
  "status_code": 200,
367
- "bytes": 5591,
368
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1",
369
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fpublic_surface_qa.json",
370
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/public_surface_qa.json"
371
  }
372
  },
@@ -385,6 +449,8 @@
385
  "mirrors": {
386
  "github_pages": {
387
  "ok": true,
 
 
388
  "status_code": 200,
389
  "bytes": 4032,
390
  "sha256": "2390c3663e9a110b9ed2f9c581b88a95bbd811a72ffb3e8dc780dae420e0f919",
@@ -393,6 +459,8 @@
393
  },
394
  "hf_space": {
395
  "ok": true,
 
 
396
  "status_code": 200,
397
  "bytes": 4032,
398
  "sha256": "2390c3663e9a110b9ed2f9c581b88a95bbd811a72ffb3e8dc780dae420e0f919",
@@ -401,18 +469,22 @@
401
  },
402
  "hf_artifacts": {
403
  "ok": true,
 
 
404
  "status_code": 200,
405
  "bytes": 4032,
406
  "sha256": "2390c3663e9a110b9ed2f9c581b88a95bbd811a72ffb3e8dc780dae420e0f919",
407
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Frendered_site_check.json",
408
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/rendered_site_check.json"
409
  },
410
  "hf_model": {
411
  "ok": true,
 
 
412
  "status_code": 200,
413
  "bytes": 4032,
414
  "sha256": "2390c3663e9a110b9ed2f9c581b88a95bbd811a72ffb3e8dc780dae420e0f919",
415
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Frendered_site_check.json",
416
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/rendered_site_check.json"
417
  }
418
  },
@@ -431,6 +503,8 @@
431
  "mirrors": {
432
  "github_pages": {
433
  "ok": true,
 
 
434
  "status_code": 200,
435
  "bytes": 7559,
436
  "sha256": "417e65746724c87cbf1cd40f6451f1f6fe6098f60ba0b1c24d11e718fc86e257",
@@ -439,6 +513,8 @@
439
  },
440
  "hf_space": {
441
  "ok": true,
 
 
442
  "status_code": 200,
443
  "bytes": 7559,
444
  "sha256": "417e65746724c87cbf1cd40f6451f1f6fe6098f60ba0b1c24d11e718fc86e257",
@@ -447,18 +523,22 @@
447
  },
448
  "hf_artifacts": {
449
  "ok": true,
 
 
450
  "status_code": 200,
451
  "bytes": 7559,
452
  "sha256": "417e65746724c87cbf1cd40f6451f1f6fe6098f60ba0b1c24d11e718fc86e257",
453
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fxperience10m_dataset_card_alignment.json",
454
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/xperience10m_dataset_card_alignment.json"
455
  },
456
  "hf_model": {
457
  "ok": true,
 
 
458
  "status_code": 200,
459
  "bytes": 7559,
460
  "sha256": "417e65746724c87cbf1cd40f6451f1f6fe6098f60ba0b1c24d11e718fc86e257",
461
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fxperience10m_dataset_card_alignment.json",
462
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/xperience10m_dataset_card_alignment.json"
463
  }
464
  },
@@ -472,39 +552,47 @@
472
  "path": "docs/data/source_alignment_audit.json",
473
  "exists": true,
474
  "bytes": 4432,
475
- "sha256": "06c6e2d111c72df01ed127fd288e6675b63e35a21ae12a2523931a072bd0bc49"
476
  },
477
  "mirrors": {
478
  "github_pages": {
479
  "ok": true,
 
 
480
  "status_code": 200,
481
  "bytes": 4432,
482
- "sha256": "06c6e2d111c72df01ed127fd288e6675b63e35a21ae12a2523931a072bd0bc49",
483
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/source_alignment_audit.json",
484
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/source_alignment_audit.json"
485
  },
486
  "hf_space": {
487
  "ok": true,
 
 
488
  "status_code": 200,
489
  "bytes": 4432,
490
- "sha256": "06c6e2d111c72df01ed127fd288e6675b63e35a21ae12a2523931a072bd0bc49",
491
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/source_alignment_audit.json",
492
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/source_alignment_audit.json"
493
  },
494
  "hf_artifacts": {
495
  "ok": true,
 
 
496
  "status_code": 200,
497
  "bytes": 4432,
498
- "sha256": "06c6e2d111c72df01ed127fd288e6675b63e35a21ae12a2523931a072bd0bc49",
499
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fsource_alignment_audit.json",
500
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/source_alignment_audit.json"
501
  },
502
  "hf_model": {
503
  "ok": true,
 
 
504
  "status_code": 200,
505
  "bytes": 4432,
506
- "sha256": "06c6e2d111c72df01ed127fd288e6675b63e35a21ae12a2523931a072bd0bc49",
507
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fsource_alignment_audit.json",
508
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/source_alignment_audit.json"
509
  }
510
  },
@@ -517,40 +605,48 @@
517
  "local": {
518
  "path": "docs/data/project_status.json",
519
  "exists": true,
520
- "bytes": 16400,
521
- "sha256": "0ef6b75a18793906e3f0c45e3b37f4023c6a3af9fcf7367bfa43924f161b6eff"
522
  },
523
  "mirrors": {
524
  "github_pages": {
525
  "ok": true,
 
 
526
  "status_code": 200,
527
- "bytes": 16400,
528
- "sha256": "0ef6b75a18793906e3f0c45e3b37f4023c6a3af9fcf7367bfa43924f161b6eff",
529
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/project_status.json",
530
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/project_status.json"
531
  },
532
  "hf_space": {
533
  "ok": true,
 
 
534
  "status_code": 200,
535
- "bytes": 16400,
536
- "sha256": "0ef6b75a18793906e3f0c45e3b37f4023c6a3af9fcf7367bfa43924f161b6eff",
537
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/project_status.json",
538
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/project_status.json"
539
  },
540
  "hf_artifacts": {
541
  "ok": true,
 
 
542
  "status_code": 200,
543
- "bytes": 16400,
544
- "sha256": "0ef6b75a18793906e3f0c45e3b37f4023c6a3af9fcf7367bfa43924f161b6eff",
545
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fproject_status.json",
546
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/project_status.json"
547
  },
548
  "hf_model": {
549
  "ok": true,
 
 
550
  "status_code": 200,
551
- "bytes": 16400,
552
- "sha256": "0ef6b75a18793906e3f0c45e3b37f4023c6a3af9fcf7367bfa43924f161b6eff",
553
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fproject_status.json",
554
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/project_status.json"
555
  }
556
  },
@@ -563,45 +659,215 @@
563
  "local": {
564
  "path": "docs/data/omni_model_comparison.json",
565
  "exists": true,
566
- "bytes": 42015,
567
- "sha256": "f74673dcc879ba73f7bc3210d8716ae5ad5d7de70ee76d59f56bb24443669c12"
568
  },
569
  "mirrors": {
570
  "github_pages": {
571
  "ok": true,
 
 
572
  "status_code": 200,
573
- "bytes": 42015,
574
- "sha256": "f74673dcc879ba73f7bc3210d8716ae5ad5d7de70ee76d59f56bb24443669c12",
575
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/omni_model_comparison.json",
576
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/omni_model_comparison.json"
577
  },
578
  "hf_space": {
579
  "ok": true,
 
 
580
  "status_code": 200,
581
- "bytes": 42015,
582
- "sha256": "f74673dcc879ba73f7bc3210d8716ae5ad5d7de70ee76d59f56bb24443669c12",
583
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/omni_model_comparison.json",
584
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/omni_model_comparison.json"
585
  },
586
  "hf_artifacts": {
587
  "ok": true,
 
 
588
  "status_code": 200,
589
- "bytes": 42015,
590
- "sha256": "f74673dcc879ba73f7bc3210d8716ae5ad5d7de70ee76d59f56bb24443669c12",
591
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fomni_model_comparison.json",
592
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/omni_model_comparison.json"
593
  },
594
  "hf_model": {
595
  "ok": true,
 
 
596
  "status_code": 200,
597
- "bytes": 42015,
598
- "sha256": "f74673dcc879ba73f7bc3210d8716ae5ad5d7de70ee76d59f56bb24443669c12",
599
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fomni_model_comparison.json",
600
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/omni_model_comparison.json"
601
  }
602
  },
603
  "failures": []
604
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
605
  {
606
  "id": "evaluation_protocol_json",
607
  "title": "Evaluation protocol JSON",
@@ -615,6 +881,8 @@
615
  "mirrors": {
616
  "github_pages": {
617
  "ok": true,
 
 
618
  "status_code": 200,
619
  "bytes": 14511,
620
  "sha256": "ea7caff963fcf048f803a852e5cdae8d3975ba4a36d805c5e42211b1bf2744ef",
@@ -623,6 +891,8 @@
623
  },
624
  "hf_space": {
625
  "ok": true,
 
 
626
  "status_code": 200,
627
  "bytes": 14511,
628
  "sha256": "ea7caff963fcf048f803a852e5cdae8d3975ba4a36d805c5e42211b1bf2744ef",
@@ -631,18 +901,22 @@
631
  },
632
  "hf_artifacts": {
633
  "ok": true,
 
 
634
  "status_code": 200,
635
  "bytes": 14511,
636
  "sha256": "ea7caff963fcf048f803a852e5cdae8d3975ba4a36d805c5e42211b1bf2744ef",
637
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fevaluation_protocol.json",
638
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/evaluation_protocol.json"
639
  },
640
  "hf_model": {
641
  "ok": true,
 
 
642
  "status_code": 200,
643
  "bytes": 14511,
644
  "sha256": "ea7caff963fcf048f803a852e5cdae8d3975ba4a36d805c5e42211b1bf2744ef",
645
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fevaluation_protocol.json",
646
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/evaluation_protocol.json"
647
  }
648
  },
@@ -661,6 +935,8 @@
661
  "mirrors": {
662
  "github_pages": {
663
  "ok": true,
 
 
664
  "status_code": 200,
665
  "bytes": 7139,
666
  "sha256": "eb87b65ef2f6ef910b4cda29c33f3c75014a5cce8ebf8299f71eb09c856a2481",
@@ -669,6 +945,8 @@
669
  },
670
  "hf_space": {
671
  "ok": true,
 
 
672
  "status_code": 200,
673
  "bytes": 7139,
674
  "sha256": "eb87b65ef2f6ef910b4cda29c33f3c75014a5cce8ebf8299f71eb09c856a2481",
@@ -677,18 +955,22 @@
677
  },
678
  "hf_artifacts": {
679
  "ok": true,
 
 
680
  "status_code": 200,
681
  "bytes": 7139,
682
  "sha256": "eb87b65ef2f6ef910b4cda29c33f3c75014a5cce8ebf8299f71eb09c856a2481",
683
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fresearch_takeaways.json",
684
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/research_takeaways.json"
685
  },
686
  "hf_model": {
687
  "ok": true,
 
 
688
  "status_code": 200,
689
  "bytes": 7139,
690
  "sha256": "eb87b65ef2f6ef910b4cda29c33f3c75014a5cce8ebf8299f71eb09c856a2481",
691
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fresearch_takeaways.json",
692
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/research_takeaways.json"
693
  }
694
  },
@@ -701,40 +983,48 @@
701
  "local": {
702
  "path": "docs/data/research_roadmap.json",
703
  "exists": true,
704
- "bytes": 10133,
705
- "sha256": "45fd3a1bde93654ccfe14f9271928a67b36eb3f166826bfbdbb9c1092ad33bcf"
706
  },
707
  "mirrors": {
708
  "github_pages": {
709
  "ok": true,
 
 
710
  "status_code": 200,
711
- "bytes": 10133,
712
- "sha256": "45fd3a1bde93654ccfe14f9271928a67b36eb3f166826bfbdbb9c1092ad33bcf",
713
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/research_roadmap.json",
714
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/research_roadmap.json"
715
  },
716
  "hf_space": {
717
  "ok": true,
 
 
718
  "status_code": 200,
719
- "bytes": 10133,
720
- "sha256": "45fd3a1bde93654ccfe14f9271928a67b36eb3f166826bfbdbb9c1092ad33bcf",
721
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/research_roadmap.json",
722
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/research_roadmap.json"
723
  },
724
  "hf_artifacts": {
725
  "ok": true,
 
 
726
  "status_code": 200,
727
- "bytes": 10133,
728
- "sha256": "45fd3a1bde93654ccfe14f9271928a67b36eb3f166826bfbdbb9c1092ad33bcf",
729
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fresearch_roadmap.json",
730
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/research_roadmap.json"
731
  },
732
  "hf_model": {
733
  "ok": true,
 
 
734
  "status_code": 200,
735
- "bytes": 10133,
736
- "sha256": "45fd3a1bde93654ccfe14f9271928a67b36eb3f166826bfbdbb9c1092ad33bcf",
737
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fresearch_roadmap.json",
738
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/research_roadmap.json"
739
  }
740
  },
@@ -747,40 +1037,48 @@
747
  "local": {
748
  "path": "docs/data/research_roadmap_interactive.json",
749
  "exists": true,
750
- "bytes": 143560,
751
- "sha256": "9198752056a40eb5a7457ded21576862d9954be1f0f4a9e996e935d328ef4062"
752
  },
753
  "mirrors": {
754
  "github_pages": {
755
  "ok": true,
 
 
756
  "status_code": 200,
757
- "bytes": 143560,
758
- "sha256": "9198752056a40eb5a7457ded21576862d9954be1f0f4a9e996e935d328ef4062",
759
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/research_roadmap_interactive.json",
760
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/research_roadmap_interactive.json"
761
  },
762
  "hf_space": {
763
  "ok": true,
 
 
764
  "status_code": 200,
765
- "bytes": 143560,
766
- "sha256": "9198752056a40eb5a7457ded21576862d9954be1f0f4a9e996e935d328ef4062",
767
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/research_roadmap_interactive.json",
768
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/research_roadmap_interactive.json"
769
  },
770
  "hf_artifacts": {
771
  "ok": true,
 
 
772
  "status_code": 200,
773
- "bytes": 143560,
774
- "sha256": "9198752056a40eb5a7457ded21576862d9954be1f0f4a9e996e935d328ef4062",
775
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fresearch_roadmap_interactive.json",
776
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/research_roadmap_interactive.json"
777
  },
778
  "hf_model": {
779
  "ok": true,
 
 
780
  "status_code": 200,
781
- "bytes": 143560,
782
- "sha256": "9198752056a40eb5a7457ded21576862d9954be1f0f4a9e996e935d328ef4062",
783
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fresearch_roadmap_interactive.json",
784
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/research_roadmap_interactive.json"
785
  }
786
  },
@@ -799,6 +1097,8 @@
799
  "mirrors": {
800
  "github_pages": {
801
  "ok": true,
 
 
802
  "status_code": 200,
803
  "bytes": 14117,
804
  "sha256": "3cb04da5c2807a3f7a5e196c5bf7d40dbe2a5d8143cefcb37323e1096889afa4",
@@ -807,6 +1107,8 @@
807
  },
808
  "hf_space": {
809
  "ok": true,
 
 
810
  "status_code": 200,
811
  "bytes": 14117,
812
  "sha256": "3cb04da5c2807a3f7a5e196c5bf7d40dbe2a5d8143cefcb37323e1096889afa4",
@@ -815,18 +1117,22 @@
815
  },
816
  "hf_artifacts": {
817
  "ok": true,
 
 
818
  "status_code": 200,
819
  "bytes": 14117,
820
  "sha256": "3cb04da5c2807a3f7a5e196c5bf7d40dbe2a5d8143cefcb37323e1096889afa4",
821
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Ffigure_index.json",
822
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/figure_index.json"
823
  },
824
  "hf_model": {
825
  "ok": true,
 
 
826
  "status_code": 200,
827
  "bytes": 14117,
828
  "sha256": "3cb04da5c2807a3f7a5e196c5bf7d40dbe2a5d8143cefcb37323e1096889afa4",
829
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Ffigure_index.json",
830
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/figure_index.json"
831
  }
832
  },
@@ -845,6 +1151,8 @@
845
  "mirrors": {
846
  "github_pages": {
847
  "ok": true,
 
 
848
  "status_code": 200,
849
  "bytes": 26963,
850
  "sha256": "b93df37e7da1fcd5d3df7c2643b4659e2fef986dba52cb37a164874919ccc53a",
@@ -853,6 +1161,8 @@
853
  },
854
  "hf_space": {
855
  "ok": true,
 
 
856
  "status_code": 200,
857
  "bytes": 26963,
858
  "sha256": "b93df37e7da1fcd5d3df7c2643b4659e2fef986dba52cb37a164874919ccc53a",
@@ -861,18 +1171,22 @@
861
  },
862
  "hf_artifacts": {
863
  "ok": true,
 
 
864
  "status_code": 200,
865
  "bytes": 26963,
866
  "sha256": "b93df37e7da1fcd5d3df7c2643b4659e2fef986dba52cb37a164874919ccc53a",
867
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Ftask_walkthroughs.json",
868
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/task_walkthroughs.json"
869
  },
870
  "hf_model": {
871
  "ok": true,
 
 
872
  "status_code": 200,
873
  "bytes": 26963,
874
  "sha256": "b93df37e7da1fcd5d3df7c2643b4659e2fef986dba52cb37a164874919ccc53a",
875
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Ftask_walkthroughs.json",
876
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/task_walkthroughs.json"
877
  }
878
  },
@@ -886,39 +1200,47 @@
886
  "path": "docs/data/task_surface_integrity.json",
887
  "exists": true,
888
  "bytes": 45779,
889
- "sha256": "b2f4fdb1384e610a8961d706db33b1ef1b25e9378d283c34d1a66e4ab0a95bab"
890
  },
891
  "mirrors": {
892
  "github_pages": {
893
  "ok": true,
 
 
894
  "status_code": 200,
895
  "bytes": 45779,
896
- "sha256": "b2f4fdb1384e610a8961d706db33b1ef1b25e9378d283c34d1a66e4ab0a95bab",
897
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/task_surface_integrity.json",
898
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/task_surface_integrity.json"
899
  },
900
  "hf_space": {
901
  "ok": true,
 
 
902
  "status_code": 200,
903
  "bytes": 45779,
904
- "sha256": "b2f4fdb1384e610a8961d706db33b1ef1b25e9378d283c34d1a66e4ab0a95bab",
905
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/task_surface_integrity.json",
906
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/task_surface_integrity.json"
907
  },
908
  "hf_artifacts": {
909
  "ok": true,
 
 
910
  "status_code": 200,
911
  "bytes": 45779,
912
- "sha256": "b2f4fdb1384e610a8961d706db33b1ef1b25e9378d283c34d1a66e4ab0a95bab",
913
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Ftask_surface_integrity.json",
914
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/task_surface_integrity.json"
915
  },
916
  "hf_model": {
917
  "ok": true,
 
 
918
  "status_code": 200,
919
  "bytes": 45779,
920
- "sha256": "b2f4fdb1384e610a8961d706db33b1ef1b25e9378d283c34d1a66e4ab0a95bab",
921
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Ftask_surface_integrity.json",
922
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/task_surface_integrity.json"
923
  }
924
  },
@@ -937,6 +1259,8 @@
937
  "mirrors": {
938
  "github_pages": {
939
  "ok": true,
 
 
940
  "status_code": 200,
941
  "bytes": 3898,
942
  "sha256": "883a4a3ec27b72bff0c1d53b72dd4ba872a4d1df0695fba5ebe3b11c036a4224",
@@ -945,6 +1269,8 @@
945
  },
946
  "hf_space": {
947
  "ok": true,
 
 
948
  "status_code": 200,
949
  "bytes": 3898,
950
  "sha256": "883a4a3ec27b72bff0c1d53b72dd4ba872a4d1df0695fba5ebe3b11c036a4224",
@@ -953,18 +1279,22 @@
953
  },
954
  "hf_artifacts": {
955
  "ok": true,
 
 
956
  "status_code": 200,
957
  "bytes": 3898,
958
  "sha256": "883a4a3ec27b72bff0c1d53b72dd4ba872a4d1df0695fba5ebe3b11c036a4224",
959
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fbrand_assets.json",
960
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/brand_assets.json"
961
  },
962
  "hf_model": {
963
  "ok": true,
 
 
964
  "status_code": 200,
965
  "bytes": 3898,
966
  "sha256": "883a4a3ec27b72bff0c1d53b72dd4ba872a4d1df0695fba5ebe3b11c036a4224",
967
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fbrand_assets.json",
968
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/brand_assets.json"
969
  }
970
  },
@@ -983,6 +1313,8 @@
983
  "mirrors": {
984
  "github_pages": {
985
  "ok": true,
 
 
986
  "status_code": 200,
987
  "bytes": 201330,
988
  "sha256": "16e99ea98eb890e02aeb681e16b5163e1d5a9702d784dbd34ce3691e423a766a",
@@ -991,14 +1323,18 @@
991
  },
992
  "hf_space": {
993
  "ok": true,
 
 
994
  "status_code": 200,
995
  "bytes": 201330,
996
  "sha256": "16e99ea98eb890e02aeb681e16b5163e1d5a9702d784dbd34ce3691e423a766a",
997
- "final_url": "https://cas-bridge.xethub.hf.co/xet-bridge-us/6a19f5db39fe6ce4ebbae226/f5463f2804fbc02b3501d8370683e1f0f8348ff3d2c7a7221a6a186da3770b66",
998
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/resolve/main/assets/brand/xperience10m-logo-social-card.png"
999
  },
1000
  "hf_artifacts": {
1001
  "ok": true,
 
 
1002
  "status_code": 200,
1003
  "bytes": 201330,
1004
  "sha256": "16e99ea98eb890e02aeb681e16b5163e1d5a9702d784dbd34ce3691e423a766a",
@@ -1007,6 +1343,8 @@
1007
  },
1008
  "hf_model": {
1009
  "ok": true,
 
 
1010
  "status_code": 200,
1011
  "bytes": 201330,
1012
  "sha256": "16e99ea98eb890e02aeb681e16b5163e1d5a9702d784dbd34ce3691e423a766a",
@@ -1029,6 +1367,8 @@
1029
  "mirrors": {
1030
  "github_raw": {
1031
  "ok": true,
 
 
1032
  "status_code": 200,
1033
  "bytes": 4880,
1034
  "sha256": "526a38edffdb8e96eb7be3fc4ae4c8fab5a43ac4ed6e57137e9e0857c75b0a27",
@@ -1037,6 +1377,8 @@
1037
  },
1038
  "hf_space": {
1039
  "ok": true,
 
 
1040
  "status_code": 200,
1041
  "bytes": 4880,
1042
  "sha256": "526a38edffdb8e96eb7be3fc4ae4c8fab5a43ac4ed6e57137e9e0857c75b0a27",
@@ -1045,6 +1387,8 @@
1045
  },
1046
  "hf_artifacts": {
1047
  "ok": true,
 
 
1048
  "status_code": 200,
1049
  "bytes": 4880,
1050
  "sha256": "526a38edffdb8e96eb7be3fc4ae4c8fab5a43ac4ed6e57137e9e0857c75b0a27",
@@ -1053,6 +1397,8 @@
1053
  },
1054
  "hf_model": {
1055
  "ok": true,
 
 
1056
  "status_code": 200,
1057
  "bytes": 4880,
1058
  "sha256": "526a38edffdb8e96eb7be3fc4ae4c8fab5a43ac4ed6e57137e9e0857c75b0a27",
@@ -1069,16 +1415,18 @@
1069
  "local": {
1070
  "path": "../hf_publish/qwen3_lora_128ep/README.md",
1071
  "exists": true,
1072
- "bytes": 3376,
1073
- "sha256": "c70fa727d670be017ce2237dddd01df5d4bf7fd9fc54927aaf6b3c01c095e43a"
1074
  },
1075
  "mirrors": {
1076
  "hf_qwen3_lora_model": {
1077
  "ok": true,
 
 
1078
  "status_code": 200,
1079
- "bytes": 3376,
1080
- "sha256": "c70fa727d670be017ce2237dddd01df5d4bf7fd9fc54927aaf6b3c01c095e43a",
1081
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-qwen3-omni-lora-128ep/7b924bec6e2c670f422d751d3d3a406880ff3219/README.md",
1082
  "url": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep/resolve/main/README.md"
1083
  }
1084
  },
@@ -1091,16 +1439,18 @@
1091
  "local": {
1092
  "path": "../hf_publish/qwen3_lora_128ep/upload_manifest.json",
1093
  "exists": true,
1094
- "bytes": 2519,
1095
- "sha256": "21c44c3eeef45fad4d5ed8c3651973cdb9c0859cd6d76c8d7984c51ba6950841"
1096
  },
1097
  "mirrors": {
1098
  "hf_qwen3_lora_model": {
1099
  "ok": true,
 
 
1100
  "status_code": 200,
1101
- "bytes": 2519,
1102
- "sha256": "21c44c3eeef45fad4d5ed8c3651973cdb9c0859cd6d76c8d7984c51ba6950841",
1103
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-qwen3-omni-lora-128ep/7b924bec6e2c670f422d751d3d3a406880ff3219/upload_manifest.json",
1104
  "url": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep/resolve/main/upload_manifest.json"
1105
  }
1106
  },
@@ -1113,16 +1463,18 @@
1113
  "local": {
1114
  "path": "../hf_publish/qwen3_lora_128ep/adapter_config.json",
1115
  "exists": true,
1116
- "bytes": 1179,
1117
- "sha256": "fc02f0a129fc8b6c0afcf7a5f29fbeba9eabc9d0990d00c9c5ece9377e248353"
1118
  },
1119
  "mirrors": {
1120
  "hf_qwen3_lora_model": {
1121
  "ok": true,
 
 
1122
  "status_code": 200,
1123
- "bytes": 1179,
1124
- "sha256": "fc02f0a129fc8b6c0afcf7a5f29fbeba9eabc9d0990d00c9c5ece9377e248353",
1125
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-qwen3-omni-lora-128ep/7b924bec6e2c670f422d751d3d3a406880ff3219/adapter_config.json",
1126
  "url": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep/resolve/main/adapter_config.json"
1127
  }
1128
  },
@@ -1136,19 +1488,189 @@
1136
  "path": "../hf_publish/qwen3_lora_128ep/adapter_model.safetensors",
1137
  "exists": true,
1138
  "bytes": 69284456,
1139
- "sha256": "97d049e1fe7e50a93e0e3179cbc175f30e75f98ec2e19a8f412347a47705eea5"
1140
  },
1141
  "mirrors": {
1142
  "hf_qwen3_lora_model": {
1143
  "ok": true,
 
 
1144
  "status_code": 200,
1145
  "bytes": 69284456,
1146
- "sha256": "97d049e1fe7e50a93e0e3179cbc175f30e75f98ec2e19a8f412347a47705eea5",
1147
- "final_url": "https://cas-bridge.xethub.hf.co/xet-bridge-us/6a24aa8f77b6bd853fd1c263/4b61a44160d713b264046d77f357992ff72e734f9a9bc365ad2074efa28e7354",
1148
  "url": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep/resolve/main/adapter_model.safetensors"
1149
  }
1150
  },
1151
  "failures": []
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1152
  }
1153
  ],
1154
  "marker_checks": [
@@ -1159,12 +1681,14 @@
1159
  "status": "pass",
1160
  "fetch": {
1161
  "ok": true,
 
 
1162
  "status_code": 200,
1163
- "bytes": 180715,
1164
- "sha256": "c1a311421a1567227e36c53a0ec0a4e993b44bff2c7ff3e43790d5fca959f7d5",
1165
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/"
1166
  },
1167
- "required_marker_count": 17,
1168
  "missing_markers": [],
1169
  "forbidden_markers_present": [],
1170
  "failures": []
@@ -1176,12 +1700,14 @@
1176
  "status": "pass",
1177
  "fetch": {
1178
  "ok": true,
 
 
1179
  "status_code": 200,
1180
- "bytes": 180715,
1181
- "sha256": "c1a311421a1567227e36c53a0ec0a4e993b44bff2c7ff3e43790d5fca959f7d5",
1182
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/index.html"
1183
  },
1184
- "required_marker_count": 17,
1185
  "missing_markers": [],
1186
  "forbidden_markers_present": [],
1187
  "failures": []
@@ -1193,12 +1719,14 @@
1193
  "status": "pass",
1194
  "fetch": {
1195
  "ok": true,
 
 
1196
  "status_code": 200,
1197
- "bytes": 3932,
1198
- "sha256": "e80c45f4d6c3c411c21597940805ee524c1ea52ff2ce81d779710f8e47d6b743",
1199
  "final_url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/raw/main/README.md"
1200
  },
1201
- "required_marker_count": 5,
1202
  "missing_markers": [],
1203
  "forbidden_markers_present": [],
1204
  "failures": []
@@ -1210,6 +1738,8 @@
1210
  "status": "pass",
1211
  "fetch": {
1212
  "ok": true,
 
 
1213
  "status_code": 200,
1214
  "bytes": 2770710,
1215
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
@@ -1227,6 +1757,8 @@
1227
  "status": "pass",
1228
  "fetch": {
1229
  "ok": true,
 
 
1230
  "status_code": 200,
1231
  "bytes": 2770710,
1232
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
@@ -1244,12 +1776,14 @@
1244
  "status": "pass",
1245
  "fetch": {
1246
  "ok": true,
 
 
1247
  "status_code": 200,
1248
- "bytes": 69045,
1249
- "sha256": "4ca4e853740529274ac3d5266b3305705582c72c97238b2e00285fdce588fcec",
1250
  "final_url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/raw/main/README.md"
1251
  },
1252
- "required_marker_count": 5,
1253
  "missing_markers": [],
1254
  "forbidden_markers_present": [],
1255
  "failures": []
@@ -1261,9 +1795,11 @@
1261
  "status": "pass",
1262
  "fetch": {
1263
  "ok": true,
 
 
1264
  "status_code": 200,
1265
- "bytes": 257049,
1266
- "sha256": "7214326d6c3a7aacd50024ba104adac84c289f02558304f7eaa8c742db3e511f",
1267
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/mirror_parity.json"
1268
  },
1269
  "required_marker_count": 1,
@@ -1278,9 +1814,11 @@
1278
  "status": "pass",
1279
  "fetch": {
1280
  "ok": true,
 
 
1281
  "status_code": 200,
1282
- "bytes": 257049,
1283
- "sha256": "7214326d6c3a7aacd50024ba104adac84c289f02558304f7eaa8c742db3e511f",
1284
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/mirror_parity.json"
1285
  },
1286
  "required_marker_count": 1,
@@ -1295,10 +1833,12 @@
1295
  "status": "pass",
1296
  "fetch": {
1297
  "ok": true,
 
 
1298
  "status_code": 200,
1299
- "bytes": 257049,
1300
- "sha256": "7214326d6c3a7aacd50024ba104adac84c289f02558304f7eaa8c742db3e511f",
1301
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fmirror_parity.json"
1302
  },
1303
  "required_marker_count": 1,
1304
  "missing_markers": [],
@@ -1312,10 +1852,12 @@
1312
  "status": "pass",
1313
  "fetch": {
1314
  "ok": true,
 
 
1315
  "status_code": 200,
1316
- "bytes": 257049,
1317
- "sha256": "7214326d6c3a7aacd50024ba104adac84c289f02558304f7eaa8c742db3e511f",
1318
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fmirror_parity.json"
1319
  },
1320
  "required_marker_count": 1,
1321
  "missing_markers": [],
@@ -1329,9 +1871,11 @@
1329
  "status": "pass",
1330
  "fetch": {
1331
  "ok": true,
 
 
1332
  "status_code": 200,
1333
- "bytes": 7212,
1334
- "sha256": "bbb46638856b3c9c5cbf752f0f9a5cbe44dfbf3f4d29633c945daafa62eb0d36",
1335
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/publication_audit.json"
1336
  },
1337
  "required_marker_count": 1,
@@ -1346,9 +1890,11 @@
1346
  "status": "pass",
1347
  "fetch": {
1348
  "ok": true,
 
 
1349
  "status_code": 200,
1350
- "bytes": 7212,
1351
- "sha256": "bbb46638856b3c9c5cbf752f0f9a5cbe44dfbf3f4d29633c945daafa62eb0d36",
1352
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/publication_audit.json"
1353
  },
1354
  "required_marker_count": 1,
@@ -1363,10 +1909,12 @@
1363
  "status": "pass",
1364
  "fetch": {
1365
  "ok": true,
 
 
1366
  "status_code": 200,
1367
- "bytes": 7212,
1368
- "sha256": "bbb46638856b3c9c5cbf752f0f9a5cbe44dfbf3f4d29633c945daafa62eb0d36",
1369
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fpublication_audit.json"
1370
  },
1371
  "required_marker_count": 1,
1372
  "missing_markers": [],
@@ -1380,10 +1928,12 @@
1380
  "status": "pass",
1381
  "fetch": {
1382
  "ok": true,
 
 
1383
  "status_code": 200,
1384
- "bytes": 7212,
1385
- "sha256": "bbb46638856b3c9c5cbf752f0f9a5cbe44dfbf3f4d29633c945daafa62eb0d36",
1386
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fpublication_audit.json"
1387
  },
1388
  "required_marker_count": 1,
1389
  "missing_markers": [],
@@ -1397,9 +1947,11 @@
1397
  "status": "pass",
1398
  "fetch": {
1399
  "ok": true,
 
 
1400
  "status_code": 200,
1401
- "bytes": 15375,
1402
- "sha256": "ee640e4a0b2271ed97b8c57ce6b23542b474c46d9baa77fdfb003c2f372b4975",
1403
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/website_integrity.json"
1404
  },
1405
  "required_marker_count": 1,
@@ -1414,9 +1966,11 @@
1414
  "status": "pass",
1415
  "fetch": {
1416
  "ok": true,
 
 
1417
  "status_code": 200,
1418
- "bytes": 15375,
1419
- "sha256": "ee640e4a0b2271ed97b8c57ce6b23542b474c46d9baa77fdfb003c2f372b4975",
1420
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/website_integrity.json"
1421
  },
1422
  "required_marker_count": 1,
@@ -1431,10 +1985,12 @@
1431
  "status": "pass",
1432
  "fetch": {
1433
  "ok": true,
 
 
1434
  "status_code": 200,
1435
- "bytes": 15375,
1436
- "sha256": "ee640e4a0b2271ed97b8c57ce6b23542b474c46d9baa77fdfb003c2f372b4975",
1437
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fwebsite_integrity.json"
1438
  },
1439
  "required_marker_count": 1,
1440
  "missing_markers": [],
@@ -1448,10 +2004,12 @@
1448
  "status": "pass",
1449
  "fetch": {
1450
  "ok": true,
 
 
1451
  "status_code": 200,
1452
- "bytes": 15375,
1453
- "sha256": "ee640e4a0b2271ed97b8c57ce6b23542b474c46d9baa77fdfb003c2f372b4975",
1454
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fwebsite_integrity.json"
1455
  },
1456
  "required_marker_count": 1,
1457
  "missing_markers": [],
@@ -1465,9 +2023,11 @@
1465
  "status": "pass",
1466
  "fetch": {
1467
  "ok": true,
 
 
1468
  "status_code": 200,
1469
- "bytes": 5591,
1470
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1",
1471
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/public_surface_qa.json"
1472
  },
1473
  "required_marker_count": 1,
@@ -1482,9 +2042,11 @@
1482
  "status": "pass",
1483
  "fetch": {
1484
  "ok": true,
 
 
1485
  "status_code": 200,
1486
- "bytes": 5591,
1487
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1",
1488
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/public_surface_qa.json"
1489
  },
1490
  "required_marker_count": 1,
@@ -1499,10 +2061,12 @@
1499
  "status": "pass",
1500
  "fetch": {
1501
  "ok": true,
 
 
1502
  "status_code": 200,
1503
- "bytes": 5591,
1504
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1",
1505
- "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/477a0b7312e42d79e4ec47803a68f1dedcc59bed/docs%2Fdata%2Fpublic_surface_qa.json"
1506
  },
1507
  "required_marker_count": 1,
1508
  "missing_markers": [],
@@ -1516,10 +2080,12 @@
1516
  "status": "pass",
1517
  "fetch": {
1518
  "ok": true,
 
 
1519
  "status_code": 200,
1520
- "bytes": 5591,
1521
- "sha256": "28f4c5fece8bfe7d46e6e8c3f5c7ca712ffe5a4ed4f5ca9e39655ada8f3e65f1",
1522
- "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/9c8c9a423b3efcc716acd36562c7cdcf0e7f9fc9/metrics%2Fpublic_surface_qa.json"
1523
  },
1524
  "required_marker_count": 1,
1525
  "missing_markers": [],
 
1
  {
2
  "title": "Ropedia Xperience-10M Live Publication Status",
3
  "status": "pass",
4
+ "checked_at_utc": "2026-06-11T04:08:43+00:00",
5
+ "scope": "Live GitHub Pages, GitHub raw, Hugging Face Space, artifact dataset, baseline model mirrors, and the Qwen3/Cosmos3 LoRA adapter repos when their upload packages exist locally.",
6
  "hash_groups": [
7
  {
8
  "id": "task_suite_infographic",
 
17
  "mirrors": {
18
  "github_pages": {
19
  "ok": true,
20
+ "method": "curl",
21
+ "timeout_seconds": 30,
22
  "status_code": 200,
23
  "bytes": 1588641,
24
  "sha256": "1275e2adaef920ecde7c29dc62c8d79d4f13475a0c09bc3baa693f47cdec2e1f",
 
27
  },
28
  "hf_space": {
29
  "ok": true,
30
+ "method": "curl",
31
+ "timeout_seconds": 30,
32
  "status_code": 200,
33
  "bytes": 1588641,
34
  "sha256": "1275e2adaef920ecde7c29dc62c8d79d4f13475a0c09bc3baa693f47cdec2e1f",
 
37
  },
38
  "hf_artifacts": {
39
  "ok": true,
40
+ "method": "curl",
41
+ "timeout_seconds": 30,
42
  "status_code": 200,
43
  "bytes": 1588641,
44
  "sha256": "1275e2adaef920ecde7c29dc62c8d79d4f13475a0c09bc3baa693f47cdec2e1f",
 
47
  },
48
  "hf_model": {
49
  "ok": true,
50
+ "method": "curl",
51
+ "timeout_seconds": 30,
52
  "status_code": 200,
53
  "bytes": 1588641,
54
  "sha256": "1275e2adaef920ecde7c29dc62c8d79d4f13475a0c09bc3baa693f47cdec2e1f",
 
65
  "local": {
66
  "path": "docs/data/quality_gates.json",
67
  "exists": true,
68
+ "bytes": 8097,
69
+ "sha256": "53ccc5522df02b2b3658958f3da6ef542a2aa9f85e0bf69cad61ca7a0d254160"
70
  },
71
  "mirrors": {
72
  "github_pages": {
73
  "ok": true,
74
+ "method": "curl",
75
+ "timeout_seconds": 30,
76
  "status_code": 200,
77
+ "bytes": 8097,
78
+ "sha256": "53ccc5522df02b2b3658958f3da6ef542a2aa9f85e0bf69cad61ca7a0d254160",
79
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/quality_gates.json",
80
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/quality_gates.json"
81
  },
82
  "hf_space": {
83
  "ok": true,
84
+ "method": "curl",
85
+ "timeout_seconds": 30,
86
  "status_code": 200,
87
+ "bytes": 8097,
88
+ "sha256": "53ccc5522df02b2b3658958f3da6ef542a2aa9f85e0bf69cad61ca7a0d254160",
89
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/quality_gates.json",
90
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/quality_gates.json"
91
  },
92
  "hf_artifacts": {
93
  "ok": true,
94
+ "method": "curl",
95
+ "timeout_seconds": 30,
96
  "status_code": 200,
97
+ "bytes": 8097,
98
+ "sha256": "53ccc5522df02b2b3658958f3da6ef542a2aa9f85e0bf69cad61ca7a0d254160",
99
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fquality_gates.json",
100
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/quality_gates.json"
101
  },
102
  "hf_model": {
103
  "ok": true,
104
+ "method": "curl",
105
+ "timeout_seconds": 30,
106
  "status_code": 200,
107
+ "bytes": 8097,
108
+ "sha256": "53ccc5522df02b2b3658958f3da6ef542a2aa9f85e0bf69cad61ca7a0d254160",
109
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fquality_gates.json",
110
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/quality_gates.json"
111
  }
112
  },
 
125
  "mirrors": {
126
  "github_pages": {
127
  "ok": true,
128
+ "method": "curl",
129
+ "timeout_seconds": 30,
130
  "status_code": 200,
131
  "bytes": 4305527,
132
  "sha256": "286230d8070150e09595797569a447308a5110241b69edf52e535e3d8a34f827",
 
135
  },
136
  "hf_space": {
137
  "ok": true,
138
+ "method": "curl",
139
+ "timeout_seconds": 30,
140
  "status_code": 200,
141
  "bytes": 4305527,
142
  "sha256": "286230d8070150e09595797569a447308a5110241b69edf52e535e3d8a34f827",
 
145
  },
146
  "hf_artifacts": {
147
  "ok": true,
148
+ "method": "curl",
149
+ "timeout_seconds": 30,
150
  "status_code": 200,
151
  "bytes": 4305527,
152
  "sha256": "286230d8070150e09595797569a447308a5110241b69edf52e535e3d8a34f827",
153
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fsingle_episode_explorer.json",
154
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/single_episode_explorer.json"
155
  },
156
  "hf_model": {
157
  "ok": true,
158
+ "method": "curl",
159
+ "timeout_seconds": 30,
160
  "status_code": 200,
161
  "bytes": 4305527,
162
  "sha256": "286230d8070150e09595797569a447308a5110241b69edf52e535e3d8a34f827",
163
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fsingle_episode_explorer.json",
164
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/single_episode_explorer.json"
165
  }
166
  },
 
179
  "mirrors": {
180
  "github_pages": {
181
  "ok": true,
182
+ "method": "curl",
183
+ "timeout_seconds": 30,
184
  "status_code": 200,
185
  "bytes": 2770710,
186
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
 
189
  },
190
  "hf_space": {
191
  "ok": true,
192
+ "method": "curl",
193
+ "timeout_seconds": 30,
194
  "status_code": 200,
195
  "bytes": 2770710,
196
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
 
199
  },
200
  "hf_artifacts": {
201
  "ok": true,
202
+ "method": "curl",
203
+ "timeout_seconds": 30,
204
  "status_code": 200,
205
  "bytes": 2770710,
206
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
207
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fsingle_episode_explorer.html",
208
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/single_episode_explorer.html"
209
  },
210
  "hf_model": {
211
  "ok": true,
212
+ "method": "curl",
213
+ "timeout_seconds": 30,
214
  "status_code": 200,
215
  "bytes": 2770710,
216
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
217
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/single_episode_explorer.html",
218
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/single_episode_explorer.html"
219
  }
220
  },
 
233
  "mirrors": {
234
  "github_pages": {
235
  "ok": true,
236
+ "method": "curl",
237
+ "timeout_seconds": 30,
238
  "status_code": 200,
239
  "bytes": 31706,
240
  "sha256": "a51ccc9e42247be6cda5a026c96be74e2aeeacfae6ef47abf028c9f769eae9a1",
 
243
  },
244
  "hf_space": {
245
  "ok": true,
246
+ "method": "curl",
247
+ "timeout_seconds": 30,
248
  "status_code": 200,
249
  "bytes": 31706,
250
  "sha256": "a51ccc9e42247be6cda5a026c96be74e2aeeacfae6ef47abf028c9f769eae9a1",
 
253
  },
254
  "hf_artifacts": {
255
  "ok": true,
256
+ "method": "curl",
257
+ "timeout_seconds": 30,
258
  "status_code": 200,
259
  "bytes": 31706,
260
  "sha256": "a51ccc9e42247be6cda5a026c96be74e2aeeacfae6ef47abf028c9f769eae9a1",
261
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fresearch_roadmap.html",
262
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/research_roadmap.html"
263
  },
264
  "hf_model": {
265
  "ok": true,
266
+ "method": "curl",
267
+ "timeout_seconds": 30,
268
  "status_code": 200,
269
  "bytes": 31706,
270
  "sha256": "a51ccc9e42247be6cda5a026c96be74e2aeeacfae6ef47abf028c9f769eae9a1",
271
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/research_roadmap.html",
272
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/research_roadmap.html"
273
  }
274
  },
 
287
  "mirrors": {
288
  "github_raw": {
289
  "ok": true,
290
+ "method": "curl",
291
+ "timeout_seconds": 30,
292
  "status_code": 200,
293
  "bytes": 4072,
294
  "sha256": "7189be7fc773943859dc75a6119c1fb654d8905ea381e8fc9f782d4b33f5e466",
 
297
  },
298
  "hf_space": {
299
  "ok": true,
300
+ "method": "curl",
301
+ "timeout_seconds": 30,
302
  "status_code": 200,
303
  "bytes": 4072,
304
  "sha256": "7189be7fc773943859dc75a6119c1fb654d8905ea381e8fc9f782d4b33f5e466",
 
307
  },
308
  "hf_artifacts": {
309
  "ok": true,
310
+ "method": "curl",
311
+ "timeout_seconds": 30,
312
  "status_code": 200,
313
  "bytes": 4072,
314
  "sha256": "7189be7fc773943859dc75a6119c1fb654d8905ea381e8fc9f782d4b33f5e466",
315
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/results%2Fsingle_episode_diagnostics%2Fprovenance.json",
316
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/results/single_episode_diagnostics/provenance.json"
317
  },
318
  "hf_model": {
319
  "ok": true,
320
+ "method": "curl",
321
+ "timeout_seconds": 30,
322
  "status_code": 200,
323
  "bytes": 4072,
324
  "sha256": "7189be7fc773943859dc75a6119c1fb654d8905ea381e8fc9f782d4b33f5e466",
325
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/results%2Fsingle_episode_diagnostics%2Fprovenance.json",
326
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/results/single_episode_diagnostics/provenance.json"
327
  }
328
  },
 
341
  "mirrors": {
342
  "github_raw": {
343
  "ok": true,
344
+ "method": "curl",
345
+ "timeout_seconds": 30,
346
  "status_code": 200,
347
  "bytes": 995,
348
  "sha256": "813320bbe5d57c74803f5dfb080d440ef1804af41950e031d82ab80311eb04c7",
 
351
  },
352
  "hf_space": {
353
  "ok": true,
354
+ "method": "curl",
355
+ "timeout_seconds": 30,
356
  "status_code": 200,
357
  "bytes": 995,
358
  "sha256": "813320bbe5d57c74803f5dfb080d440ef1804af41950e031d82ab80311eb04c7",
 
361
  },
362
  "hf_artifacts": {
363
  "ok": true,
364
+ "method": "curl",
365
+ "timeout_seconds": 30,
366
  "status_code": 200,
367
  "bytes": 995,
368
  "sha256": "813320bbe5d57c74803f5dfb080d440ef1804af41950e031d82ab80311eb04c7",
369
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/results%2Fsingle_episode_diagnostics%2Fobject_labels%2Fobject_vocab.json",
370
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/results/single_episode_diagnostics/object_labels/object_vocab.json"
371
  },
372
  "hf_model": {
373
  "ok": true,
374
+ "method": "curl",
375
+ "timeout_seconds": 30,
376
  "status_code": 200,
377
  "bytes": 995,
378
  "sha256": "813320bbe5d57c74803f5dfb080d440ef1804af41950e031d82ab80311eb04c7",
379
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/results%2Fsingle_episode_diagnostics%2Fobject_labels%2Fobject_vocab.json",
380
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/results/single_episode_diagnostics/object_labels/object_vocab.json"
381
  }
382
  },
 
389
  "local": {
390
  "path": "docs/data/public_surface_qa.json",
391
  "exists": true,
392
+ "bytes": 5640,
393
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2"
394
  },
395
  "mirrors": {
396
  "github_pages": {
397
  "ok": true,
398
+ "method": "curl",
399
+ "timeout_seconds": 30,
400
  "status_code": 200,
401
+ "bytes": 5640,
402
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2",
403
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/public_surface_qa.json",
404
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/public_surface_qa.json"
405
  },
406
  "hf_space": {
407
  "ok": true,
408
+ "method": "curl",
409
+ "timeout_seconds": 30,
410
  "status_code": 200,
411
+ "bytes": 5640,
412
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2",
413
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/public_surface_qa.json",
414
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/public_surface_qa.json"
415
  },
416
  "hf_artifacts": {
417
  "ok": true,
418
+ "method": "curl",
419
+ "timeout_seconds": 30,
420
  "status_code": 200,
421
+ "bytes": 5640,
422
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2",
423
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fpublic_surface_qa.json",
424
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/public_surface_qa.json"
425
  },
426
  "hf_model": {
427
  "ok": true,
428
+ "method": "curl",
429
+ "timeout_seconds": 30,
430
  "status_code": 200,
431
+ "bytes": 5640,
432
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2",
433
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fpublic_surface_qa.json",
434
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/public_surface_qa.json"
435
  }
436
  },
 
449
  "mirrors": {
450
  "github_pages": {
451
  "ok": true,
452
+ "method": "curl",
453
+ "timeout_seconds": 30,
454
  "status_code": 200,
455
  "bytes": 4032,
456
  "sha256": "2390c3663e9a110b9ed2f9c581b88a95bbd811a72ffb3e8dc780dae420e0f919",
 
459
  },
460
  "hf_space": {
461
  "ok": true,
462
+ "method": "curl",
463
+ "timeout_seconds": 30,
464
  "status_code": 200,
465
  "bytes": 4032,
466
  "sha256": "2390c3663e9a110b9ed2f9c581b88a95bbd811a72ffb3e8dc780dae420e0f919",
 
469
  },
470
  "hf_artifacts": {
471
  "ok": true,
472
+ "method": "curl",
473
+ "timeout_seconds": 30,
474
  "status_code": 200,
475
  "bytes": 4032,
476
  "sha256": "2390c3663e9a110b9ed2f9c581b88a95bbd811a72ffb3e8dc780dae420e0f919",
477
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Frendered_site_check.json",
478
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/rendered_site_check.json"
479
  },
480
  "hf_model": {
481
  "ok": true,
482
+ "method": "curl",
483
+ "timeout_seconds": 30,
484
  "status_code": 200,
485
  "bytes": 4032,
486
  "sha256": "2390c3663e9a110b9ed2f9c581b88a95bbd811a72ffb3e8dc780dae420e0f919",
487
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Frendered_site_check.json",
488
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/rendered_site_check.json"
489
  }
490
  },
 
503
  "mirrors": {
504
  "github_pages": {
505
  "ok": true,
506
+ "method": "curl",
507
+ "timeout_seconds": 30,
508
  "status_code": 200,
509
  "bytes": 7559,
510
  "sha256": "417e65746724c87cbf1cd40f6451f1f6fe6098f60ba0b1c24d11e718fc86e257",
 
513
  },
514
  "hf_space": {
515
  "ok": true,
516
+ "method": "curl",
517
+ "timeout_seconds": 30,
518
  "status_code": 200,
519
  "bytes": 7559,
520
  "sha256": "417e65746724c87cbf1cd40f6451f1f6fe6098f60ba0b1c24d11e718fc86e257",
 
523
  },
524
  "hf_artifacts": {
525
  "ok": true,
526
+ "method": "curl",
527
+ "timeout_seconds": 30,
528
  "status_code": 200,
529
  "bytes": 7559,
530
  "sha256": "417e65746724c87cbf1cd40f6451f1f6fe6098f60ba0b1c24d11e718fc86e257",
531
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fxperience10m_dataset_card_alignment.json",
532
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/xperience10m_dataset_card_alignment.json"
533
  },
534
  "hf_model": {
535
  "ok": true,
536
+ "method": "curl",
537
+ "timeout_seconds": 30,
538
  "status_code": 200,
539
  "bytes": 7559,
540
  "sha256": "417e65746724c87cbf1cd40f6451f1f6fe6098f60ba0b1c24d11e718fc86e257",
541
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fxperience10m_dataset_card_alignment.json",
542
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/xperience10m_dataset_card_alignment.json"
543
  }
544
  },
 
552
  "path": "docs/data/source_alignment_audit.json",
553
  "exists": true,
554
  "bytes": 4432,
555
+ "sha256": "01d976b900fdbbccd580346a2e68eafe1784ed889ec105803a43fb768a8c9064"
556
  },
557
  "mirrors": {
558
  "github_pages": {
559
  "ok": true,
560
+ "method": "curl",
561
+ "timeout_seconds": 30,
562
  "status_code": 200,
563
  "bytes": 4432,
564
+ "sha256": "01d976b900fdbbccd580346a2e68eafe1784ed889ec105803a43fb768a8c9064",
565
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/source_alignment_audit.json",
566
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/source_alignment_audit.json"
567
  },
568
  "hf_space": {
569
  "ok": true,
570
+ "method": "curl",
571
+ "timeout_seconds": 30,
572
  "status_code": 200,
573
  "bytes": 4432,
574
+ "sha256": "01d976b900fdbbccd580346a2e68eafe1784ed889ec105803a43fb768a8c9064",
575
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/source_alignment_audit.json",
576
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/source_alignment_audit.json"
577
  },
578
  "hf_artifacts": {
579
  "ok": true,
580
+ "method": "curl",
581
+ "timeout_seconds": 30,
582
  "status_code": 200,
583
  "bytes": 4432,
584
+ "sha256": "01d976b900fdbbccd580346a2e68eafe1784ed889ec105803a43fb768a8c9064",
585
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fsource_alignment_audit.json",
586
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/source_alignment_audit.json"
587
  },
588
  "hf_model": {
589
  "ok": true,
590
+ "method": "curl",
591
+ "timeout_seconds": 30,
592
  "status_code": 200,
593
  "bytes": 4432,
594
+ "sha256": "01d976b900fdbbccd580346a2e68eafe1784ed889ec105803a43fb768a8c9064",
595
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fsource_alignment_audit.json",
596
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/source_alignment_audit.json"
597
  }
598
  },
 
605
  "local": {
606
  "path": "docs/data/project_status.json",
607
  "exists": true,
608
+ "bytes": 23198,
609
+ "sha256": "07678354fde90a7e15134f5deff036eb200f7bf422ab2350ff64825b902843d5"
610
  },
611
  "mirrors": {
612
  "github_pages": {
613
  "ok": true,
614
+ "method": "curl",
615
+ "timeout_seconds": 30,
616
  "status_code": 200,
617
+ "bytes": 23198,
618
+ "sha256": "07678354fde90a7e15134f5deff036eb200f7bf422ab2350ff64825b902843d5",
619
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/project_status.json",
620
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/project_status.json"
621
  },
622
  "hf_space": {
623
  "ok": true,
624
+ "method": "curl",
625
+ "timeout_seconds": 30,
626
  "status_code": 200,
627
+ "bytes": 23198,
628
+ "sha256": "07678354fde90a7e15134f5deff036eb200f7bf422ab2350ff64825b902843d5",
629
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/project_status.json",
630
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/project_status.json"
631
  },
632
  "hf_artifacts": {
633
  "ok": true,
634
+ "method": "curl",
635
+ "timeout_seconds": 30,
636
  "status_code": 200,
637
+ "bytes": 23198,
638
+ "sha256": "07678354fde90a7e15134f5deff036eb200f7bf422ab2350ff64825b902843d5",
639
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fproject_status.json",
640
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/project_status.json"
641
  },
642
  "hf_model": {
643
  "ok": true,
644
+ "method": "curl",
645
+ "timeout_seconds": 30,
646
  "status_code": 200,
647
+ "bytes": 23198,
648
+ "sha256": "07678354fde90a7e15134f5deff036eb200f7bf422ab2350ff64825b902843d5",
649
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fproject_status.json",
650
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/project_status.json"
651
  }
652
  },
 
659
  "local": {
660
  "path": "docs/data/omni_model_comparison.json",
661
  "exists": true,
662
+ "bytes": 75698,
663
+ "sha256": "3b310ccf4e1c72aba1382f051778d581fd150cef468db826869ade8e8149abca"
664
  },
665
  "mirrors": {
666
  "github_pages": {
667
  "ok": true,
668
+ "method": "curl",
669
+ "timeout_seconds": 30,
670
  "status_code": 200,
671
+ "bytes": 75698,
672
+ "sha256": "3b310ccf4e1c72aba1382f051778d581fd150cef468db826869ade8e8149abca",
673
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/omni_model_comparison.json",
674
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/omni_model_comparison.json"
675
  },
676
  "hf_space": {
677
  "ok": true,
678
+ "method": "curl",
679
+ "timeout_seconds": 30,
680
  "status_code": 200,
681
+ "bytes": 75698,
682
+ "sha256": "3b310ccf4e1c72aba1382f051778d581fd150cef468db826869ade8e8149abca",
683
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/omni_model_comparison.json",
684
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/omni_model_comparison.json"
685
  },
686
  "hf_artifacts": {
687
  "ok": true,
688
+ "method": "curl",
689
+ "timeout_seconds": 30,
690
  "status_code": 200,
691
+ "bytes": 75698,
692
+ "sha256": "3b310ccf4e1c72aba1382f051778d581fd150cef468db826869ade8e8149abca",
693
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fomni_model_comparison.json",
694
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/omni_model_comparison.json"
695
  },
696
  "hf_model": {
697
  "ok": true,
698
+ "method": "curl",
699
+ "timeout_seconds": 30,
700
  "status_code": 200,
701
+ "bytes": 75698,
702
+ "sha256": "3b310ccf4e1c72aba1382f051778d581fd150cef468db826869ade8e8149abca",
703
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fomni_model_comparison.json",
704
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/omni_model_comparison.json"
705
  }
706
  },
707
  "failures": []
708
  },
709
+ {
710
+ "id": "task_suite_enhancement_128_json",
711
+ "title": "128-episode task-suite enhancement JSON",
712
+ "status": "pass",
713
+ "local": {
714
+ "path": "docs/data/task_suite_enhancement_128.json",
715
+ "exists": true,
716
+ "bytes": 20181,
717
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e"
718
+ },
719
+ "mirrors": {
720
+ "github_pages": {
721
+ "ok": true,
722
+ "method": "curl",
723
+ "timeout_seconds": 30,
724
+ "status_code": 200,
725
+ "bytes": 20181,
726
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e",
727
+ "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/task_suite_enhancement_128.json",
728
+ "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/task_suite_enhancement_128.json"
729
+ },
730
+ "hf_space": {
731
+ "ok": true,
732
+ "method": "curl",
733
+ "timeout_seconds": 30,
734
+ "status_code": 200,
735
+ "bytes": 20181,
736
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e",
737
+ "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/task_suite_enhancement_128.json",
738
+ "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/task_suite_enhancement_128.json"
739
+ },
740
+ "hf_artifacts": {
741
+ "ok": true,
742
+ "method": "curl",
743
+ "timeout_seconds": 30,
744
+ "status_code": 200,
745
+ "bytes": 20181,
746
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e",
747
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Ftask_suite_enhancement_128.json",
748
+ "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/task_suite_enhancement_128.json"
749
+ },
750
+ "hf_model": {
751
+ "ok": true,
752
+ "method": "curl",
753
+ "timeout_seconds": 30,
754
+ "status_code": 200,
755
+ "bytes": 20181,
756
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e",
757
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Ftask_suite_enhancement_128.json",
758
+ "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/task_suite_enhancement_128.json"
759
+ }
760
+ },
761
+ "failures": []
762
+ },
763
+ {
764
+ "id": "task_suite_enhancement_128_markdown",
765
+ "title": "128-episode task-suite enhancement Markdown",
766
+ "status": "pass",
767
+ "local": {
768
+ "path": "TASK_SUITE_ENHANCEMENT_128.md",
769
+ "exists": true,
770
+ "bytes": 3427,
771
+ "sha256": "f6b30c2c288ed2dd2b1b1bb36c334d9697665201c60f77374d38b1ff3911fc86"
772
+ },
773
+ "mirrors": {
774
+ "github_raw": {
775
+ "ok": true,
776
+ "method": "curl",
777
+ "timeout_seconds": 30,
778
+ "status_code": 200,
779
+ "bytes": 3427,
780
+ "sha256": "f6b30c2c288ed2dd2b1b1bb36c334d9697665201c60f77374d38b1ff3911fc86",
781
+ "final_url": "https://raw.githubusercontent.com/ChaoYue0307/ropedia-xperience-10m-task-suite/main/TASK_SUITE_ENHANCEMENT_128.md",
782
+ "url": "https://raw.githubusercontent.com/ChaoYue0307/ropedia-xperience-10m-task-suite/main/TASK_SUITE_ENHANCEMENT_128.md"
783
+ },
784
+ "hf_space": {
785
+ "ok": true,
786
+ "method": "curl",
787
+ "timeout_seconds": 30,
788
+ "status_code": 200,
789
+ "bytes": 3427,
790
+ "sha256": "f6b30c2c288ed2dd2b1b1bb36c334d9697665201c60f77374d38b1ff3911fc86",
791
+ "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/TASK_SUITE_ENHANCEMENT_128.md",
792
+ "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/TASK_SUITE_ENHANCEMENT_128.md"
793
+ },
794
+ "hf_artifacts": {
795
+ "ok": true,
796
+ "method": "curl",
797
+ "timeout_seconds": 30,
798
+ "status_code": 200,
799
+ "bytes": 3427,
800
+ "sha256": "f6b30c2c288ed2dd2b1b1bb36c334d9697665201c60f77374d38b1ff3911fc86",
801
+ "final_url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/raw/main/TASK_SUITE_ENHANCEMENT_128.md",
802
+ "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/raw/main/TASK_SUITE_ENHANCEMENT_128.md"
803
+ },
804
+ "hf_model": {
805
+ "ok": true,
806
+ "method": "curl",
807
+ "timeout_seconds": 30,
808
+ "status_code": 200,
809
+ "bytes": 3427,
810
+ "sha256": "f6b30c2c288ed2dd2b1b1bb36c334d9697665201c60f77374d38b1ff3911fc86",
811
+ "final_url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/raw/main/TASK_SUITE_ENHANCEMENT_128.md",
812
+ "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/raw/main/TASK_SUITE_ENHANCEMENT_128.md"
813
+ }
814
+ },
815
+ "failures": []
816
+ },
817
+ {
818
+ "id": "task_suite_enhancement_128_plan",
819
+ "title": "128-episode task-suite enhancement plan",
820
+ "status": "pass",
821
+ "local": {
822
+ "path": "results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json",
823
+ "exists": true,
824
+ "bytes": 20181,
825
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e"
826
+ },
827
+ "mirrors": {
828
+ "github_raw": {
829
+ "ok": true,
830
+ "method": "curl",
831
+ "timeout_seconds": 30,
832
+ "status_code": 200,
833
+ "bytes": 20181,
834
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e",
835
+ "final_url": "https://raw.githubusercontent.com/ChaoYue0307/ropedia-xperience-10m-task-suite/main/results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json",
836
+ "url": "https://raw.githubusercontent.com/ChaoYue0307/ropedia-xperience-10m-task-suite/main/results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json"
837
+ },
838
+ "hf_space": {
839
+ "ok": true,
840
+ "method": "curl",
841
+ "timeout_seconds": 30,
842
+ "status_code": 200,
843
+ "bytes": 20181,
844
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e",
845
+ "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json",
846
+ "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json"
847
+ },
848
+ "hf_artifacts": {
849
+ "ok": true,
850
+ "method": "curl",
851
+ "timeout_seconds": 30,
852
+ "status_code": 200,
853
+ "bytes": 20181,
854
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e",
855
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/results%2Fomni_finetune%2Ftask_suite_enhancement_128_v1_20260608%2Fenhancement_plan.json",
856
+ "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json"
857
+ },
858
+ "hf_model": {
859
+ "ok": true,
860
+ "method": "curl",
861
+ "timeout_seconds": 30,
862
+ "status_code": 200,
863
+ "bytes": 20181,
864
+ "sha256": "17453f9a949278b1f3038d68124f8f0e2441584d4c1384d482d2ac9ca295e97e",
865
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/results%2Fomni_finetune%2Ftask_suite_enhancement_128_v1_20260608%2Fenhancement_plan.json",
866
+ "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json"
867
+ }
868
+ },
869
+ "failures": []
870
+ },
871
  {
872
  "id": "evaluation_protocol_json",
873
  "title": "Evaluation protocol JSON",
 
881
  "mirrors": {
882
  "github_pages": {
883
  "ok": true,
884
+ "method": "curl",
885
+ "timeout_seconds": 30,
886
  "status_code": 200,
887
  "bytes": 14511,
888
  "sha256": "ea7caff963fcf048f803a852e5cdae8d3975ba4a36d805c5e42211b1bf2744ef",
 
891
  },
892
  "hf_space": {
893
  "ok": true,
894
+ "method": "curl",
895
+ "timeout_seconds": 30,
896
  "status_code": 200,
897
  "bytes": 14511,
898
  "sha256": "ea7caff963fcf048f803a852e5cdae8d3975ba4a36d805c5e42211b1bf2744ef",
 
901
  },
902
  "hf_artifacts": {
903
  "ok": true,
904
+ "method": "curl",
905
+ "timeout_seconds": 30,
906
  "status_code": 200,
907
  "bytes": 14511,
908
  "sha256": "ea7caff963fcf048f803a852e5cdae8d3975ba4a36d805c5e42211b1bf2744ef",
909
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fevaluation_protocol.json",
910
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/evaluation_protocol.json"
911
  },
912
  "hf_model": {
913
  "ok": true,
914
+ "method": "curl",
915
+ "timeout_seconds": 30,
916
  "status_code": 200,
917
  "bytes": 14511,
918
  "sha256": "ea7caff963fcf048f803a852e5cdae8d3975ba4a36d805c5e42211b1bf2744ef",
919
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fevaluation_protocol.json",
920
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/evaluation_protocol.json"
921
  }
922
  },
 
935
  "mirrors": {
936
  "github_pages": {
937
  "ok": true,
938
+ "method": "curl",
939
+ "timeout_seconds": 30,
940
  "status_code": 200,
941
  "bytes": 7139,
942
  "sha256": "eb87b65ef2f6ef910b4cda29c33f3c75014a5cce8ebf8299f71eb09c856a2481",
 
945
  },
946
  "hf_space": {
947
  "ok": true,
948
+ "method": "curl",
949
+ "timeout_seconds": 30,
950
  "status_code": 200,
951
  "bytes": 7139,
952
  "sha256": "eb87b65ef2f6ef910b4cda29c33f3c75014a5cce8ebf8299f71eb09c856a2481",
 
955
  },
956
  "hf_artifacts": {
957
  "ok": true,
958
+ "method": "curl",
959
+ "timeout_seconds": 30,
960
  "status_code": 200,
961
  "bytes": 7139,
962
  "sha256": "eb87b65ef2f6ef910b4cda29c33f3c75014a5cce8ebf8299f71eb09c856a2481",
963
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fresearch_takeaways.json",
964
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/research_takeaways.json"
965
  },
966
  "hf_model": {
967
  "ok": true,
968
+ "method": "curl",
969
+ "timeout_seconds": 30,
970
  "status_code": 200,
971
  "bytes": 7139,
972
  "sha256": "eb87b65ef2f6ef910b4cda29c33f3c75014a5cce8ebf8299f71eb09c856a2481",
973
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fresearch_takeaways.json",
974
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/research_takeaways.json"
975
  }
976
  },
 
983
  "local": {
984
  "path": "docs/data/research_roadmap.json",
985
  "exists": true,
986
+ "bytes": 13313,
987
+ "sha256": "719a8b5f45810d6b09788391d8549eb5226e480454c9f2a4ed7380ceacfd834b"
988
  },
989
  "mirrors": {
990
  "github_pages": {
991
  "ok": true,
992
+ "method": "curl",
993
+ "timeout_seconds": 30,
994
  "status_code": 200,
995
+ "bytes": 13313,
996
+ "sha256": "719a8b5f45810d6b09788391d8549eb5226e480454c9f2a4ed7380ceacfd834b",
997
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/research_roadmap.json",
998
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/research_roadmap.json"
999
  },
1000
  "hf_space": {
1001
  "ok": true,
1002
+ "method": "curl",
1003
+ "timeout_seconds": 30,
1004
  "status_code": 200,
1005
+ "bytes": 13313,
1006
+ "sha256": "719a8b5f45810d6b09788391d8549eb5226e480454c9f2a4ed7380ceacfd834b",
1007
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/research_roadmap.json",
1008
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/research_roadmap.json"
1009
  },
1010
  "hf_artifacts": {
1011
  "ok": true,
1012
+ "method": "curl",
1013
+ "timeout_seconds": 30,
1014
  "status_code": 200,
1015
+ "bytes": 13313,
1016
+ "sha256": "719a8b5f45810d6b09788391d8549eb5226e480454c9f2a4ed7380ceacfd834b",
1017
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fresearch_roadmap.json",
1018
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/research_roadmap.json"
1019
  },
1020
  "hf_model": {
1021
  "ok": true,
1022
+ "method": "curl",
1023
+ "timeout_seconds": 30,
1024
  "status_code": 200,
1025
+ "bytes": 13313,
1026
+ "sha256": "719a8b5f45810d6b09788391d8549eb5226e480454c9f2a4ed7380ceacfd834b",
1027
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fresearch_roadmap.json",
1028
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/research_roadmap.json"
1029
  }
1030
  },
 
1037
  "local": {
1038
  "path": "docs/data/research_roadmap_interactive.json",
1039
  "exists": true,
1040
+ "bytes": 145083,
1041
+ "sha256": "3f3896f728d65efdee792a255a878243a7f1d777f3c28aac680a6755741e8e6f"
1042
  },
1043
  "mirrors": {
1044
  "github_pages": {
1045
  "ok": true,
1046
+ "method": "curl",
1047
+ "timeout_seconds": 30,
1048
  "status_code": 200,
1049
+ "bytes": 145083,
1050
+ "sha256": "3f3896f728d65efdee792a255a878243a7f1d777f3c28aac680a6755741e8e6f",
1051
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/research_roadmap_interactive.json",
1052
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/research_roadmap_interactive.json"
1053
  },
1054
  "hf_space": {
1055
  "ok": true,
1056
+ "method": "curl",
1057
+ "timeout_seconds": 30,
1058
  "status_code": 200,
1059
+ "bytes": 145083,
1060
+ "sha256": "3f3896f728d65efdee792a255a878243a7f1d777f3c28aac680a6755741e8e6f",
1061
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/research_roadmap_interactive.json",
1062
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/research_roadmap_interactive.json"
1063
  },
1064
  "hf_artifacts": {
1065
  "ok": true,
1066
+ "method": "curl",
1067
+ "timeout_seconds": 30,
1068
  "status_code": 200,
1069
+ "bytes": 145083,
1070
+ "sha256": "3f3896f728d65efdee792a255a878243a7f1d777f3c28aac680a6755741e8e6f",
1071
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fresearch_roadmap_interactive.json",
1072
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/research_roadmap_interactive.json"
1073
  },
1074
  "hf_model": {
1075
  "ok": true,
1076
+ "method": "curl",
1077
+ "timeout_seconds": 30,
1078
  "status_code": 200,
1079
+ "bytes": 145083,
1080
+ "sha256": "3f3896f728d65efdee792a255a878243a7f1d777f3c28aac680a6755741e8e6f",
1081
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fresearch_roadmap_interactive.json",
1082
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/research_roadmap_interactive.json"
1083
  }
1084
  },
 
1097
  "mirrors": {
1098
  "github_pages": {
1099
  "ok": true,
1100
+ "method": "curl",
1101
+ "timeout_seconds": 30,
1102
  "status_code": 200,
1103
  "bytes": 14117,
1104
  "sha256": "3cb04da5c2807a3f7a5e196c5bf7d40dbe2a5d8143cefcb37323e1096889afa4",
 
1107
  },
1108
  "hf_space": {
1109
  "ok": true,
1110
+ "method": "curl",
1111
+ "timeout_seconds": 30,
1112
  "status_code": 200,
1113
  "bytes": 14117,
1114
  "sha256": "3cb04da5c2807a3f7a5e196c5bf7d40dbe2a5d8143cefcb37323e1096889afa4",
 
1117
  },
1118
  "hf_artifacts": {
1119
  "ok": true,
1120
+ "method": "curl",
1121
+ "timeout_seconds": 30,
1122
  "status_code": 200,
1123
  "bytes": 14117,
1124
  "sha256": "3cb04da5c2807a3f7a5e196c5bf7d40dbe2a5d8143cefcb37323e1096889afa4",
1125
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Ffigure_index.json",
1126
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/figure_index.json"
1127
  },
1128
  "hf_model": {
1129
  "ok": true,
1130
+ "method": "curl",
1131
+ "timeout_seconds": 30,
1132
  "status_code": 200,
1133
  "bytes": 14117,
1134
  "sha256": "3cb04da5c2807a3f7a5e196c5bf7d40dbe2a5d8143cefcb37323e1096889afa4",
1135
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Ffigure_index.json",
1136
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/figure_index.json"
1137
  }
1138
  },
 
1151
  "mirrors": {
1152
  "github_pages": {
1153
  "ok": true,
1154
+ "method": "curl",
1155
+ "timeout_seconds": 30,
1156
  "status_code": 200,
1157
  "bytes": 26963,
1158
  "sha256": "b93df37e7da1fcd5d3df7c2643b4659e2fef986dba52cb37a164874919ccc53a",
 
1161
  },
1162
  "hf_space": {
1163
  "ok": true,
1164
+ "method": "curl",
1165
+ "timeout_seconds": 30,
1166
  "status_code": 200,
1167
  "bytes": 26963,
1168
  "sha256": "b93df37e7da1fcd5d3df7c2643b4659e2fef986dba52cb37a164874919ccc53a",
 
1171
  },
1172
  "hf_artifacts": {
1173
  "ok": true,
1174
+ "method": "curl",
1175
+ "timeout_seconds": 30,
1176
  "status_code": 200,
1177
  "bytes": 26963,
1178
  "sha256": "b93df37e7da1fcd5d3df7c2643b4659e2fef986dba52cb37a164874919ccc53a",
1179
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Ftask_walkthroughs.json",
1180
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/task_walkthroughs.json"
1181
  },
1182
  "hf_model": {
1183
  "ok": true,
1184
+ "method": "curl",
1185
+ "timeout_seconds": 30,
1186
  "status_code": 200,
1187
  "bytes": 26963,
1188
  "sha256": "b93df37e7da1fcd5d3df7c2643b4659e2fef986dba52cb37a164874919ccc53a",
1189
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Ftask_walkthroughs.json",
1190
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/task_walkthroughs.json"
1191
  }
1192
  },
 
1200
  "path": "docs/data/task_surface_integrity.json",
1201
  "exists": true,
1202
  "bytes": 45779,
1203
+ "sha256": "0dca3369fd95afdbc89c9907170523209fbfc2ca127ad439d25c709c63251f8c"
1204
  },
1205
  "mirrors": {
1206
  "github_pages": {
1207
  "ok": true,
1208
+ "method": "curl",
1209
+ "timeout_seconds": 30,
1210
  "status_code": 200,
1211
  "bytes": 45779,
1212
+ "sha256": "0dca3369fd95afdbc89c9907170523209fbfc2ca127ad439d25c709c63251f8c",
1213
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/task_surface_integrity.json",
1214
  "url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/task_surface_integrity.json"
1215
  },
1216
  "hf_space": {
1217
  "ok": true,
1218
+ "method": "curl",
1219
+ "timeout_seconds": 30,
1220
  "status_code": 200,
1221
  "bytes": 45779,
1222
+ "sha256": "0dca3369fd95afdbc89c9907170523209fbfc2ca127ad439d25c709c63251f8c",
1223
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/task_surface_integrity.json",
1224
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/task_surface_integrity.json"
1225
  },
1226
  "hf_artifacts": {
1227
  "ok": true,
1228
+ "method": "curl",
1229
+ "timeout_seconds": 30,
1230
  "status_code": 200,
1231
  "bytes": 45779,
1232
+ "sha256": "0dca3369fd95afdbc89c9907170523209fbfc2ca127ad439d25c709c63251f8c",
1233
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Ftask_surface_integrity.json",
1234
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/task_surface_integrity.json"
1235
  },
1236
  "hf_model": {
1237
  "ok": true,
1238
+ "method": "curl",
1239
+ "timeout_seconds": 30,
1240
  "status_code": 200,
1241
  "bytes": 45779,
1242
+ "sha256": "0dca3369fd95afdbc89c9907170523209fbfc2ca127ad439d25c709c63251f8c",
1243
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Ftask_surface_integrity.json",
1244
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/task_surface_integrity.json"
1245
  }
1246
  },
 
1259
  "mirrors": {
1260
  "github_pages": {
1261
  "ok": true,
1262
+ "method": "curl",
1263
+ "timeout_seconds": 30,
1264
  "status_code": 200,
1265
  "bytes": 3898,
1266
  "sha256": "883a4a3ec27b72bff0c1d53b72dd4ba872a4d1df0695fba5ebe3b11c036a4224",
 
1269
  },
1270
  "hf_space": {
1271
  "ok": true,
1272
+ "method": "curl",
1273
+ "timeout_seconds": 30,
1274
  "status_code": 200,
1275
  "bytes": 3898,
1276
  "sha256": "883a4a3ec27b72bff0c1d53b72dd4ba872a4d1df0695fba5ebe3b11c036a4224",
 
1279
  },
1280
  "hf_artifacts": {
1281
  "ok": true,
1282
+ "method": "curl",
1283
+ "timeout_seconds": 30,
1284
  "status_code": 200,
1285
  "bytes": 3898,
1286
  "sha256": "883a4a3ec27b72bff0c1d53b72dd4ba872a4d1df0695fba5ebe3b11c036a4224",
1287
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fbrand_assets.json",
1288
  "url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/resolve/main/docs/data/brand_assets.json"
1289
  },
1290
  "hf_model": {
1291
  "ok": true,
1292
+ "method": "curl",
1293
+ "timeout_seconds": 30,
1294
  "status_code": 200,
1295
  "bytes": 3898,
1296
  "sha256": "883a4a3ec27b72bff0c1d53b72dd4ba872a4d1df0695fba5ebe3b11c036a4224",
1297
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fbrand_assets.json",
1298
  "url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/resolve/main/metrics/brand_assets.json"
1299
  }
1300
  },
 
1313
  "mirrors": {
1314
  "github_pages": {
1315
  "ok": true,
1316
+ "method": "curl",
1317
+ "timeout_seconds": 30,
1318
  "status_code": 200,
1319
  "bytes": 201330,
1320
  "sha256": "16e99ea98eb890e02aeb681e16b5163e1d5a9702d784dbd34ce3691e423a766a",
 
1323
  },
1324
  "hf_space": {
1325
  "ok": true,
1326
+ "method": "curl",
1327
+ "timeout_seconds": 30,
1328
  "status_code": 200,
1329
  "bytes": 201330,
1330
  "sha256": "16e99ea98eb890e02aeb681e16b5163e1d5a9702d784dbd34ce3691e423a766a",
1331
+ "final_url": "https://us.aws.cdn.hf.co/xet-bridge-us/6a19f5db39fe6ce4ebbae226/f5463f2804fbc02b3501d8370683e1f0f8348ff3d2c7a7221a6a186da3770b66",
1332
  "url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/resolve/main/assets/brand/xperience10m-logo-social-card.png"
1333
  },
1334
  "hf_artifacts": {
1335
  "ok": true,
1336
+ "method": "curl",
1337
+ "timeout_seconds": 30,
1338
  "status_code": 200,
1339
  "bytes": 201330,
1340
  "sha256": "16e99ea98eb890e02aeb681e16b5163e1d5a9702d784dbd34ce3691e423a766a",
 
1343
  },
1344
  "hf_model": {
1345
  "ok": true,
1346
+ "method": "curl",
1347
+ "timeout_seconds": 30,
1348
  "status_code": 200,
1349
  "bytes": 201330,
1350
  "sha256": "16e99ea98eb890e02aeb681e16b5163e1d5a9702d784dbd34ce3691e423a766a",
 
1367
  "mirrors": {
1368
  "github_raw": {
1369
  "ok": true,
1370
+ "method": "curl",
1371
+ "timeout_seconds": 30,
1372
  "status_code": 200,
1373
  "bytes": 4880,
1374
  "sha256": "526a38edffdb8e96eb7be3fc4ae4c8fab5a43ac4ed6e57137e9e0857c75b0a27",
 
1377
  },
1378
  "hf_space": {
1379
  "ok": true,
1380
+ "method": "curl",
1381
+ "timeout_seconds": 30,
1382
  "status_code": 200,
1383
  "bytes": 4880,
1384
  "sha256": "526a38edffdb8e96eb7be3fc4ae4c8fab5a43ac4ed6e57137e9e0857c75b0a27",
 
1387
  },
1388
  "hf_artifacts": {
1389
  "ok": true,
1390
+ "method": "curl",
1391
+ "timeout_seconds": 30,
1392
  "status_code": 200,
1393
  "bytes": 4880,
1394
  "sha256": "526a38edffdb8e96eb7be3fc4ae4c8fab5a43ac4ed6e57137e9e0857c75b0a27",
 
1397
  },
1398
  "hf_model": {
1399
  "ok": true,
1400
+ "method": "curl",
1401
+ "timeout_seconds": 30,
1402
  "status_code": 200,
1403
  "bytes": 4880,
1404
  "sha256": "526a38edffdb8e96eb7be3fc4ae4c8fab5a43ac4ed6e57137e9e0857c75b0a27",
 
1415
  "local": {
1416
  "path": "../hf_publish/qwen3_lora_128ep/README.md",
1417
  "exists": true,
1418
+ "bytes": 3366,
1419
+ "sha256": "a54c2e064d055d416589249773d6ccace038284593cc2863d24021a0fd68e510"
1420
  },
1421
  "mirrors": {
1422
  "hf_qwen3_lora_model": {
1423
  "ok": true,
1424
+ "method": "curl",
1425
+ "timeout_seconds": 30,
1426
  "status_code": 200,
1427
+ "bytes": 3366,
1428
+ "sha256": "a54c2e064d055d416589249773d6ccace038284593cc2863d24021a0fd68e510",
1429
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-qwen3-omni-lora-128ep/63cb285c582d6dda6fdac362bcb8d30a6706abbd/README.md",
1430
  "url": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep/resolve/main/README.md"
1431
  }
1432
  },
 
1439
  "local": {
1440
  "path": "../hf_publish/qwen3_lora_128ep/upload_manifest.json",
1441
  "exists": true,
1442
+ "bytes": 2457,
1443
+ "sha256": "5868d36e1581c0af4235ce172c20892ec8eedfc68ec7cb06ea65125d5997597f"
1444
  },
1445
  "mirrors": {
1446
  "hf_qwen3_lora_model": {
1447
  "ok": true,
1448
+ "method": "curl",
1449
+ "timeout_seconds": 30,
1450
  "status_code": 200,
1451
+ "bytes": 2457,
1452
+ "sha256": "5868d36e1581c0af4235ce172c20892ec8eedfc68ec7cb06ea65125d5997597f",
1453
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-qwen3-omni-lora-128ep/63cb285c582d6dda6fdac362bcb8d30a6706abbd/upload_manifest.json",
1454
  "url": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep/resolve/main/upload_manifest.json"
1455
  }
1456
  },
 
1463
  "local": {
1464
  "path": "../hf_publish/qwen3_lora_128ep/adapter_config.json",
1465
  "exists": true,
1466
+ "bytes": 1187,
1467
+ "sha256": "0ab4588c2f80843904211f4b544fdf905e5908d1b5414c6db2a704eb23659307"
1468
  },
1469
  "mirrors": {
1470
  "hf_qwen3_lora_model": {
1471
  "ok": true,
1472
+ "method": "curl",
1473
+ "timeout_seconds": 30,
1474
  "status_code": 200,
1475
+ "bytes": 1187,
1476
+ "sha256": "0ab4588c2f80843904211f4b544fdf905e5908d1b5414c6db2a704eb23659307",
1477
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-qwen3-omni-lora-128ep/63cb285c582d6dda6fdac362bcb8d30a6706abbd/adapter_config.json",
1478
  "url": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep/resolve/main/adapter_config.json"
1479
  }
1480
  },
 
1488
  "path": "../hf_publish/qwen3_lora_128ep/adapter_model.safetensors",
1489
  "exists": true,
1490
  "bytes": 69284456,
1491
+ "sha256": "1e2d8e1039dc6d658ea3b9d12d2009eef8539691abab1e65717d83be6cb0315a"
1492
  },
1493
  "mirrors": {
1494
  "hf_qwen3_lora_model": {
1495
  "ok": true,
1496
+ "method": "curl",
1497
+ "timeout_seconds": 240,
1498
  "status_code": 200,
1499
  "bytes": 69284456,
1500
+ "sha256": "1e2d8e1039dc6d658ea3b9d12d2009eef8539691abab1e65717d83be6cb0315a",
1501
+ "final_url": "https://cas-bridge.xethub.hf.co/xet-bridge-us/6a24aa8f77b6bd853fd1c263/236f594683af69eddac3fc4aeedcf4f33892444dd9caa87ab082496ddaf56c3e",
1502
  "url": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep/resolve/main/adapter_model.safetensors"
1503
  }
1504
  },
1505
  "failures": []
1506
+ },
1507
+ {
1508
+ "id": "cosmos3_super_lora_README_md",
1509
+ "title": "Cosmos3-Super LoRA repo file: README.md",
1510
+ "status": "pass",
1511
+ "local": {
1512
+ "path": "../hf_publish/cosmos3_super_forward_dynamics_lora_128ep/README.md",
1513
+ "exists": true,
1514
+ "bytes": 3379,
1515
+ "sha256": "96fe569772530be5b86431e3c79344f4c37720d16ee4a1b727e6ad16ee1a04a4"
1516
+ },
1517
+ "mirrors": {
1518
+ "hf_cosmos3_super_lora_model": {
1519
+ "ok": true,
1520
+ "method": "curl",
1521
+ "timeout_seconds": 30,
1522
+ "status_code": 200,
1523
+ "bytes": 3379,
1524
+ "sha256": "96fe569772530be5b86431e3c79344f4c37720d16ee4a1b727e6ad16ee1a04a4",
1525
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/a50ddad07d5f8bfe57d6218cad79cc65b295d90c/README.md",
1526
+ "url": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/resolve/main/README.md"
1527
+ }
1528
+ },
1529
+ "failures": []
1530
+ },
1531
+ {
1532
+ "id": "cosmos3_super_lora_upload_manifest_json",
1533
+ "title": "Cosmos3-Super LoRA repo file: upload_manifest.json",
1534
+ "status": "pass",
1535
+ "local": {
1536
+ "path": "../hf_publish/cosmos3_super_forward_dynamics_lora_128ep/upload_manifest.json",
1537
+ "exists": true,
1538
+ "bytes": 2295,
1539
+ "sha256": "4d73067b9f69ed13aa3e1ad8de47b33823de6c79d7bfa053ea32d3ccc9799276"
1540
+ },
1541
+ "mirrors": {
1542
+ "hf_cosmos3_super_lora_model": {
1543
+ "ok": true,
1544
+ "method": "curl",
1545
+ "timeout_seconds": 30,
1546
+ "status_code": 200,
1547
+ "bytes": 2295,
1548
+ "sha256": "4d73067b9f69ed13aa3e1ad8de47b33823de6c79d7bfa053ea32d3ccc9799276",
1549
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/a50ddad07d5f8bfe57d6218cad79cc65b295d90c/upload_manifest.json",
1550
+ "url": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/resolve/main/upload_manifest.json"
1551
+ }
1552
+ },
1553
+ "failures": []
1554
+ },
1555
+ {
1556
+ "id": "cosmos3_super_lora_pytorch_lora_weights_safetensors",
1557
+ "title": "Cosmos3-Super LoRA repo file: pytorch_lora_weights.safetensors",
1558
+ "status": "pass",
1559
+ "local": {
1560
+ "path": "../hf_publish/cosmos3_super_forward_dynamics_lora_128ep/pytorch_lora_weights.safetensors",
1561
+ "exists": true,
1562
+ "bytes": 104903208,
1563
+ "sha256": "44dd7614e7b7986a07bfeff4c6e897c4fbec7406b84e52c5e4b321cdf85ced16"
1564
+ },
1565
+ "mirrors": {
1566
+ "hf_cosmos3_super_lora_model": {
1567
+ "ok": true,
1568
+ "method": "curl",
1569
+ "timeout_seconds": 240,
1570
+ "status_code": 200,
1571
+ "bytes": 104903208,
1572
+ "sha256": "44dd7614e7b7986a07bfeff4c6e897c4fbec7406b84e52c5e4b321cdf85ced16",
1573
+ "final_url": "https://cas-bridge.xethub.hf.co/xet-bridge-us/6a266de81103266fee961edb/931735664032375a541323af6bdc60b1d8f8dffef7542f7a9ccdec34e661d034",
1574
+ "url": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/resolve/main/pytorch_lora_weights.safetensors"
1575
+ }
1576
+ },
1577
+ "failures": []
1578
+ },
1579
+ {
1580
+ "id": "cosmos3_super_lora_verified_result_summary_json",
1581
+ "title": "Cosmos3-Super LoRA repo file: verified_result_summary.json",
1582
+ "status": "pass",
1583
+ "local": {
1584
+ "path": "../hf_publish/cosmos3_super_forward_dynamics_lora_128ep/verified_result_summary.json",
1585
+ "exists": true,
1586
+ "bytes": 3978,
1587
+ "sha256": "f70d7507537abcf9ac140f0f1b44d0dad7c5b324cc35635fd70c8f6f85d4ae94"
1588
+ },
1589
+ "mirrors": {
1590
+ "hf_cosmos3_super_lora_model": {
1591
+ "ok": true,
1592
+ "method": "curl",
1593
+ "timeout_seconds": 30,
1594
+ "status_code": 200,
1595
+ "bytes": 3978,
1596
+ "sha256": "f70d7507537abcf9ac140f0f1b44d0dad7c5b324cc35635fd70c8f6f85d4ae94",
1597
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/a50ddad07d5f8bfe57d6218cad79cc65b295d90c/verified_result_summary.json",
1598
+ "url": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/resolve/main/verified_result_summary.json"
1599
+ }
1600
+ },
1601
+ "failures": []
1602
+ },
1603
+ {
1604
+ "id": "cosmos3_super_lora_package_audit_json",
1605
+ "title": "Cosmos3-Super LoRA repo file: package_audit.json",
1606
+ "status": "pass",
1607
+ "local": {
1608
+ "path": "../hf_publish/cosmos3_super_forward_dynamics_lora_128ep/package_audit.json",
1609
+ "exists": true,
1610
+ "bytes": 534,
1611
+ "sha256": "9d63e5f0febe52f1dd78a67b3ef4db10b3ecc40c2ddd704b73b03995b220e35e"
1612
+ },
1613
+ "mirrors": {
1614
+ "hf_cosmos3_super_lora_model": {
1615
+ "ok": true,
1616
+ "method": "curl",
1617
+ "timeout_seconds": 30,
1618
+ "status_code": 200,
1619
+ "bytes": 534,
1620
+ "sha256": "9d63e5f0febe52f1dd78a67b3ef4db10b3ecc40c2ddd704b73b03995b220e35e",
1621
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/a50ddad07d5f8bfe57d6218cad79cc65b295d90c/package_audit.json",
1622
+ "url": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/resolve/main/package_audit.json"
1623
+ }
1624
+ },
1625
+ "failures": []
1626
+ },
1627
+ {
1628
+ "id": "cosmos3_super_lora_eval_metrics_json",
1629
+ "title": "Cosmos3-Super LoRA repo file: eval_metrics.json",
1630
+ "status": "pass",
1631
+ "local": {
1632
+ "path": "../hf_publish/cosmos3_super_forward_dynamics_lora_128ep/eval_metrics.json",
1633
+ "exists": true,
1634
+ "bytes": 142983,
1635
+ "sha256": "40c0863faead8e0a852fd1efb3ba6f9e97cc8dd8327dd0d3a9fd12f9fff3b615"
1636
+ },
1637
+ "mirrors": {
1638
+ "hf_cosmos3_super_lora_model": {
1639
+ "ok": true,
1640
+ "method": "curl",
1641
+ "timeout_seconds": 30,
1642
+ "status_code": 200,
1643
+ "bytes": 142983,
1644
+ "sha256": "40c0863faead8e0a852fd1efb3ba6f9e97cc8dd8327dd0d3a9fd12f9fff3b615",
1645
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/a50ddad07d5f8bfe57d6218cad79cc65b295d90c/eval_metrics.json",
1646
+ "url": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/resolve/main/eval_metrics.json"
1647
+ }
1648
+ },
1649
+ "failures": []
1650
+ },
1651
+ {
1652
+ "id": "cosmos3_super_lora_val_metrics_json",
1653
+ "title": "Cosmos3-Super LoRA repo file: val_metrics.json",
1654
+ "status": "pass",
1655
+ "local": {
1656
+ "path": "../hf_publish/cosmos3_super_forward_dynamics_lora_128ep/val_metrics.json",
1657
+ "exists": true,
1658
+ "bytes": 146067,
1659
+ "sha256": "0b7b86b2778fc62082fb55ed3a124c934353211692fdde8f2e52408614979b40"
1660
+ },
1661
+ "mirrors": {
1662
+ "hf_cosmos3_super_lora_model": {
1663
+ "ok": true,
1664
+ "method": "curl",
1665
+ "timeout_seconds": 30,
1666
+ "status_code": 200,
1667
+ "bytes": 146067,
1668
+ "sha256": "0b7b86b2778fc62082fb55ed3a124c934353211692fdde8f2e52408614979b40",
1669
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/a50ddad07d5f8bfe57d6218cad79cc65b295d90c/val_metrics.json",
1670
+ "url": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep/resolve/main/val_metrics.json"
1671
+ }
1672
+ },
1673
+ "failures": []
1674
  }
1675
  ],
1676
  "marker_checks": [
 
1681
  "status": "pass",
1682
  "fetch": {
1683
  "ok": true,
1684
+ "method": "curl",
1685
+ "timeout_seconds": 30,
1686
  "status_code": 200,
1687
+ "bytes": 186490,
1688
+ "sha256": "f0d74ebaf983f1d49643069efe505cd45a39731349af85329b89fe8cbc5b1758",
1689
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/"
1690
  },
1691
+ "required_marker_count": 21,
1692
  "missing_markers": [],
1693
  "forbidden_markers_present": [],
1694
  "failures": []
 
1700
  "status": "pass",
1701
  "fetch": {
1702
  "ok": true,
1703
+ "method": "curl",
1704
+ "timeout_seconds": 30,
1705
  "status_code": 200,
1706
+ "bytes": 186490,
1707
+ "sha256": "f0d74ebaf983f1d49643069efe505cd45a39731349af85329b89fe8cbc5b1758",
1708
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/index.html"
1709
  },
1710
+ "required_marker_count": 21,
1711
  "missing_markers": [],
1712
  "forbidden_markers_present": [],
1713
  "failures": []
 
1719
  "status": "pass",
1720
  "fetch": {
1721
  "ok": true,
1722
+ "method": "curl",
1723
+ "timeout_seconds": 30,
1724
  "status_code": 200,
1725
+ "bytes": 5217,
1726
+ "sha256": "e05612cbe6516bdd16c10459e5d28ed9e0cc9272027d6ab3ff86e2a4dcbc5700",
1727
  "final_url": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/raw/main/README.md"
1728
  },
1729
+ "required_marker_count": 7,
1730
  "missing_markers": [],
1731
  "forbidden_markers_present": [],
1732
  "failures": []
 
1738
  "status": "pass",
1739
  "fetch": {
1740
  "ok": true,
1741
+ "method": "curl",
1742
+ "timeout_seconds": 30,
1743
  "status_code": 200,
1744
  "bytes": 2770710,
1745
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
 
1757
  "status": "pass",
1758
  "fetch": {
1759
  "ok": true,
1760
+ "method": "curl",
1761
+ "timeout_seconds": 30,
1762
  "status_code": 200,
1763
  "bytes": 2770710,
1764
  "sha256": "fea187795e701d5cf1b3d0cfd0ddb7d2c407a85309238a15006d4f928029213a",
 
1776
  "status": "pass",
1777
  "fetch": {
1778
  "ok": true,
1779
+ "method": "curl",
1780
+ "timeout_seconds": 30,
1781
  "status_code": 200,
1782
+ "bytes": 71291,
1783
+ "sha256": "aa8566178b22b11f361497f053a6241ce7f4d5b262a421963f4df1bab94074ae",
1784
  "final_url": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines/raw/main/README.md"
1785
  },
1786
+ "required_marker_count": 7,
1787
  "missing_markers": [],
1788
  "forbidden_markers_present": [],
1789
  "failures": []
 
1795
  "status": "pass",
1796
  "fetch": {
1797
  "ok": true,
1798
+ "method": "curl",
1799
+ "timeout_seconds": 30,
1800
  "status_code": 200,
1801
+ "bytes": 495491,
1802
+ "sha256": "2c4e5e77e5c9462df668cda4fc3e8eb8fae4c74bada15362114f965f3bacd694",
1803
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/mirror_parity.json"
1804
  },
1805
  "required_marker_count": 1,
 
1814
  "status": "pass",
1815
  "fetch": {
1816
  "ok": true,
1817
+ "method": "curl",
1818
+ "timeout_seconds": 30,
1819
  "status_code": 200,
1820
+ "bytes": 321031,
1821
+ "sha256": "fc1c5969252b3a7fa6e7ae2cadd48d9e9684b97d1066738eb29ff2ba96d8a3c3",
1822
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/mirror_parity.json"
1823
  },
1824
  "required_marker_count": 1,
 
1833
  "status": "pass",
1834
  "fetch": {
1835
  "ok": true,
1836
+ "method": "curl",
1837
+ "timeout_seconds": 30,
1838
  "status_code": 200,
1839
+ "bytes": 321031,
1840
+ "sha256": "fc1c5969252b3a7fa6e7ae2cadd48d9e9684b97d1066738eb29ff2ba96d8a3c3",
1841
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fmirror_parity.json"
1842
  },
1843
  "required_marker_count": 1,
1844
  "missing_markers": [],
 
1852
  "status": "pass",
1853
  "fetch": {
1854
  "ok": true,
1855
+ "method": "curl",
1856
+ "timeout_seconds": 30,
1857
  "status_code": 200,
1858
+ "bytes": 321031,
1859
+ "sha256": "fc1c5969252b3a7fa6e7ae2cadd48d9e9684b97d1066738eb29ff2ba96d8a3c3",
1860
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fmirror_parity.json"
1861
  },
1862
  "required_marker_count": 1,
1863
  "missing_markers": [],
 
1871
  "status": "pass",
1872
  "fetch": {
1873
  "ok": true,
1874
+ "method": "curl",
1875
+ "timeout_seconds": 30,
1876
  "status_code": 200,
1877
+ "bytes": 7566,
1878
+ "sha256": "c4b2aea7e63c8b05778ef9b0111895cef1afb9ff57d90b7bb85a100cf221b5b9",
1879
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/publication_audit.json"
1880
  },
1881
  "required_marker_count": 1,
 
1890
  "status": "pass",
1891
  "fetch": {
1892
  "ok": true,
1893
+ "method": "curl",
1894
+ "timeout_seconds": 30,
1895
  "status_code": 200,
1896
+ "bytes": 7566,
1897
+ "sha256": "c4b2aea7e63c8b05778ef9b0111895cef1afb9ff57d90b7bb85a100cf221b5b9",
1898
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/publication_audit.json"
1899
  },
1900
  "required_marker_count": 1,
 
1909
  "status": "pass",
1910
  "fetch": {
1911
  "ok": true,
1912
+ "method": "curl",
1913
+ "timeout_seconds": 30,
1914
  "status_code": 200,
1915
+ "bytes": 7566,
1916
+ "sha256": "c4b2aea7e63c8b05778ef9b0111895cef1afb9ff57d90b7bb85a100cf221b5b9",
1917
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fpublication_audit.json"
1918
  },
1919
  "required_marker_count": 1,
1920
  "missing_markers": [],
 
1928
  "status": "pass",
1929
  "fetch": {
1930
  "ok": true,
1931
+ "method": "curl",
1932
+ "timeout_seconds": 30,
1933
  "status_code": 200,
1934
+ "bytes": 7566,
1935
+ "sha256": "c4b2aea7e63c8b05778ef9b0111895cef1afb9ff57d90b7bb85a100cf221b5b9",
1936
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fpublication_audit.json"
1937
  },
1938
  "required_marker_count": 1,
1939
  "missing_markers": [],
 
1947
  "status": "pass",
1948
  "fetch": {
1949
  "ok": true,
1950
+ "method": "curl",
1951
+ "timeout_seconds": 30,
1952
  "status_code": 200,
1953
+ "bytes": 15654,
1954
+ "sha256": "baa3fc9d6dd6174f31fdb8217255ffe13033d3371f17ab056aaca84c6f538232",
1955
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/website_integrity.json"
1956
  },
1957
  "required_marker_count": 1,
 
1966
  "status": "pass",
1967
  "fetch": {
1968
  "ok": true,
1969
+ "method": "curl",
1970
+ "timeout_seconds": 30,
1971
  "status_code": 200,
1972
+ "bytes": 15654,
1973
+ "sha256": "baa3fc9d6dd6174f31fdb8217255ffe13033d3371f17ab056aaca84c6f538232",
1974
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/website_integrity.json"
1975
  },
1976
  "required_marker_count": 1,
 
1985
  "status": "pass",
1986
  "fetch": {
1987
  "ok": true,
1988
+ "method": "curl",
1989
+ "timeout_seconds": 30,
1990
  "status_code": 200,
1991
+ "bytes": 15654,
1992
+ "sha256": "baa3fc9d6dd6174f31fdb8217255ffe13033d3371f17ab056aaca84c6f538232",
1993
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fwebsite_integrity.json"
1994
  },
1995
  "required_marker_count": 1,
1996
  "missing_markers": [],
 
2004
  "status": "pass",
2005
  "fetch": {
2006
  "ok": true,
2007
+ "method": "curl",
2008
+ "timeout_seconds": 30,
2009
  "status_code": 200,
2010
+ "bytes": 15654,
2011
+ "sha256": "baa3fc9d6dd6174f31fdb8217255ffe13033d3371f17ab056aaca84c6f538232",
2012
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fwebsite_integrity.json"
2013
  },
2014
  "required_marker_count": 1,
2015
  "missing_markers": [],
 
2023
  "status": "pass",
2024
  "fetch": {
2025
  "ok": true,
2026
+ "method": "curl",
2027
+ "timeout_seconds": 30,
2028
  "status_code": 200,
2029
+ "bytes": 5640,
2030
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2",
2031
  "final_url": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/data/public_surface_qa.json"
2032
  },
2033
  "required_marker_count": 1,
 
2042
  "status": "pass",
2043
  "fetch": {
2044
  "ok": true,
2045
+ "method": "curl",
2046
+ "timeout_seconds": 30,
2047
  "status_code": 200,
2048
+ "bytes": 5640,
2049
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2",
2050
  "final_url": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite/raw/main/data/public_surface_qa.json"
2051
  },
2052
  "required_marker_count": 1,
 
2061
  "status": "pass",
2062
  "fetch": {
2063
  "ok": true,
2064
+ "method": "curl",
2065
+ "timeout_seconds": 30,
2066
  "status_code": 200,
2067
+ "bytes": 5640,
2068
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2",
2069
+ "final_url": "https://huggingface.co/api/resolve-cache/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts/d78e80dd584565431e2fca8f434207c1196e0ee1/docs%2Fdata%2Fpublic_surface_qa.json"
2070
  },
2071
  "required_marker_count": 1,
2072
  "missing_markers": [],
 
2080
  "status": "pass",
2081
  "fetch": {
2082
  "ok": true,
2083
+ "method": "curl",
2084
+ "timeout_seconds": 30,
2085
  "status_code": 200,
2086
+ "bytes": 5640,
2087
+ "sha256": "8a21a53e39ee5c9bed98238d15dd070da2b55c7a29b0a599c05095706b2217f2",
2088
+ "final_url": "https://huggingface.co/api/resolve-cache/models/cy0307/ropedia-xperience-10m-task-baselines/d10ad0148b25637370d587280e52d4e26c38a560/metrics%2Fpublic_surface_qa.json"
2089
  },
2090
  "required_marker_count": 1,
2091
  "missing_markers": [],
data/mirror_parity.json CHANGED
The diff for this file is too large to render. See raw diff
 
data/omni_finetune_verified_result.json CHANGED
@@ -1,86 +1,98 @@
1
  {
2
- "title": "Verified Qwen3-Omni LoRA 128-Episode Held-Out Result",
3
- "status": "verified_full_128_episode_diagnostic_result",
4
- "status_date": "2026-06-07",
5
- "backbone": "Qwen/Qwen3-Omni-30B-A3B-Instruct",
6
- "adapter": "Qwen3-Omni LoRA",
7
- "dataset": "Ropedia Xperience-10M selected 128-episode pilot",
8
- "split_policy": {
9
- "unit": "episode",
10
- "selected_episode_counts": {
11
- "train": 96,
12
- "val": 16,
13
- "test": 16
 
 
 
 
 
 
 
 
 
 
 
 
 
14
  },
15
- "exported_window_counts": {
16
- "train": 2848,
17
- "val": 512,
18
- "test": 448
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
19
  },
20
- "exported_episode_counts": {
21
- "train": 89,
22
- "val": 16,
23
- "test": 14
 
 
 
 
 
 
 
 
 
 
 
 
 
24
  },
25
- "skipped_selected_episodes": 9,
26
- "leakage_policy": "Train, validation, and test are separated by episode/session; test windows are used only for held-out evaluation."
27
- },
28
- "training": {
29
- "num_processes": 8,
30
- "epochs": 2,
31
- "lora_rank": 16,
32
- "lora_alpha": 32,
33
- "lora_dropout": 0.05,
34
- "num_train_samples": 2848,
35
- "num_val_samples": 512,
36
- "history": [
37
- {
38
- "epoch": 1,
39
- "train_loss": 0.41282760031950355,
40
- "val_loss": 0.03288277983665466,
41
- "global_step": 356
42
- },
43
- {
44
- "epoch": 2,
45
- "train_loss": 0.027745448225544075,
46
- "val_loss": 0.027823254466056824,
47
- "global_step": 712
48
- }
49
- ],
50
- "loss": "answer-token cross entropy over supervised JSON tokens",
51
- "note": "This current Qwen3-Omni LoRA result reuses the selected 96/16/16 episode setup and the v2 trained adapter, then applies the stricter label-contract prompt for held-out evaluation."
52
- },
53
- "evaluation": {
54
- "split": "test",
55
- "num_samples": 448,
56
- "held_out_episode_count": 14,
57
- "json_validity_rate": 1.0,
58
- "action_macro_f1": 0.0021983997167007384,
59
- "subtask_accuracy": 0.002232142857142857,
60
- "transition_accuracy": 0.9732142857142857,
61
- "next_action_accuracy": 0.03125,
62
- "contact_accuracy": 0.7209821428571429,
63
- "object_micro_f1": 0.30688228657389993,
64
- "quality_target": {
65
- "json_validity_rate": 0.98,
66
- "status": "met"
67
  },
68
- "previous_validation_aware_json_validity_rate": 0.875,
69
- "previous_structured_json_v2_json_validity_rate": 0.9977678571428571
70
- },
71
- "interpretation": "This is the current verified Qwen3-Omni LoRA diagnostic result for the selected 128-episode setup. It reuses the same trained LoRA adapter as v2 but tightens the prompt-side label contract at evaluation time, reaching 100% JSON validity and small gains in transition, contact, next-action exact accuracy, and object micro-F1. Action and subtask classification remain weak on held-out episodes, so this is still a baseline-quality diagnostic model rather than a strong Xperience-10M action recognizer.",
72
- "public_package": {
73
- "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v3_strict_label_prompt_reuse_lora_eval_test_full",
74
- "audit_status": "pass",
75
- "contains_raw_xperience10m_data": false,
76
- "contains_qwen_base_weights": false,
77
- "contains_lora_weights": false,
78
- "adapter_weights_repo": "cy0307/ropedia-qwen3-omni-lora-128ep"
79
- },
80
- "required_next_steps": [
81
- "Use the v3 strict-label predictions for action/subtask error analysis and unseen-label debugging.",
82
- "Keep the existing Qwen LoRA adapter repository as the weight-bearing artifact; v3 is an evaluation/package refresh over the same adapter, not new weights.",
83
- "Implement the Cosmos3-Super pipeline-loaded batch packer and one-sample forward-dynamics overfit before claiming Cosmos3 fine-tuning; camera-pose proxy targets are now exported, contract-audited, and schema-packed, but no Cosmos weights have been updated.",
84
- "Use sharded Qwen eval for future long held-out passes to improve GPU utilization."
85
- ]
86
  }
 
1
  {
2
+ "title": "Verified Qwen3-Omni LoRA 128-Episode Held-Out Result",
3
+ "status": "verified_full_128_episode_diagnostic_result",
4
+ "status_date": "2026-06-08",
5
+ "backbone": "Qwen/Qwen3-Omni-30B-A3B-Instruct",
6
+ "adapter": "Qwen3-Omni LoRA",
7
+ "dataset": "Ropedia Xperience-10M selected 128-episode pilot",
8
+ "split_policy": {
9
+ "unit": "episode",
10
+ "selected_episode_counts": {
11
+ "test": 16,
12
+ "train": 96,
13
+ "val": 16
14
+ },
15
+ "exported_window_counts": {
16
+ "train": 2848,
17
+ "val": 512,
18
+ "test": 448
19
+ },
20
+ "exported_episode_counts": {
21
+ "train": 89,
22
+ "val": 16,
23
+ "test": 14
24
+ },
25
+ "skipped_selected_episodes": 9,
26
+ "leakage_policy": "Train, validation, and test are separated by episode/session; test windows are used only for held-out evaluation."
27
  },
28
+ "training": {
29
+ "num_processes": 8,
30
+ "epochs": 4,
31
+ "lora_rank": 16,
32
+ "lora_alpha": 32,
33
+ "lora_dropout": 0.05,
34
+ "num_train_samples": 2848,
35
+ "num_val_samples": 512,
36
+ "history": [
37
+ {
38
+ "epoch": 1,
39
+ "train_loss": 0.40796751019628613,
40
+ "val_loss": 0.03258896619081497,
41
+ "global_step": 356
42
+ },
43
+ {
44
+ "epoch": 2,
45
+ "train_loss": 0.027628723937453012,
46
+ "val_loss": 0.027754632756114006,
47
+ "global_step": 712
48
+ },
49
+ {
50
+ "epoch": 3,
51
+ "train_loss": 0.02446955946807781,
52
+ "val_loss": 0.026343274861574173,
53
+ "global_step": 1068
54
+ },
55
+ {
56
+ "epoch": 4,
57
+ "train_loss": 0.022728607045444712,
58
+ "val_loss": 0.025629229843616486,
59
+ "global_step": 1424
60
+ }
61
+ ],
62
+ "loss": "answer-token cross entropy over supervised JSON tokens",
63
+ "note": "This current Qwen3-Omni LoRA result is the v4 four-epoch full held-out evaluation on the selected 96/16/16 episode setup."
64
  },
65
+ "evaluation": {
66
+ "split": "test",
67
+ "num_samples": 448,
68
+ "held_out_episode_count": 14,
69
+ "json_validity_rate": 1.0,
70
+ "action_macro_f1": 0.0018678269676001454,
71
+ "subtask_accuracy": 0.0,
72
+ "transition_accuracy": 0.9732142857142857,
73
+ "next_action_accuracy": 0.033482142857142856,
74
+ "contact_accuracy": 0.7299107142857143,
75
+ "object_micro_f1": 0.31099781500364165,
76
+ "quality_target": {
77
+ "json_validity_rate": 0.98,
78
+ "status": "met"
79
+ },
80
+ "previous_strict_label_v3_action_macro_f1": 0.0021983997167007384,
81
+ "previous_structured_json_v2_json_validity_rate": 0.9977678571428571
82
  },
83
+ "interpretation": "This is the current verified Qwen3-Omni LoRA diagnostic result for the selected 128-episode setup. The v4 four-epoch package reaches 100% JSON validity and slightly improves next-action, contact, and object metrics versus the prior strict-label v3 package, while action and subtask classification remain weak on held-out episodes. Treat it as a diagnostic baseline and error-analysis source, not as a strong Xperience-10M action recognizer.",
84
+ "public_package": {
85
+ "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
86
+ "audit_status": "pass",
87
+ "contains_raw_xperience10m_data": false,
88
+ "contains_qwen_base_weights": false,
89
+ "contains_lora_weights": false,
90
+ "adapter_weights_repo": "cy0307/ropedia-qwen3-omni-lora-128ep"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
91
  },
92
+ "required_next_steps": [
93
+ "Use the v4 predictions for action/subtask error analysis, unseen-label debugging, and hierarchical action-family scoring.",
94
+ "Use TASK_SUITE_ENHANCEMENT_128.md and docs/data/task_suite_enhancement_128.json for the no-new-episode suite push before requesting more storage.",
95
+ "Keep the existing Qwen LoRA adapter repository as the weight-bearing artifact and publish future Qwen v5 runs as separate verified packages.",
96
+ "Use the verified Cosmos3-Super Forward-Dynamics LoRA package as a separate world-model branch: it updates adapter weights over camera-pose proxy future-vision-velocity targets, not Qwen-style JSON action labels."
97
+ ]
 
 
 
 
 
 
 
 
 
 
 
 
98
  }
data/omni_model_comparison.json CHANGED
@@ -1,14 +1,14 @@
1
  {
2
  "title": "Ropedia Xperience-10M Current Result Versions and Model Groups",
3
- "generated_at_utc": "2026-06-07T17:27:36+00:00",
4
  "status": "pass",
5
  "version_count": 3,
6
- "model_group_count": 4,
7
  "comparison_rule": "Compare only rows with the same scope and target. Single-episode raw-feature metrics, 128-episode metadata baselines, Qwen3 structured JSON metrics, and the two Cosmos3 targets answer different questions: Nano future-window retrieval versus Super structured JSON Reasoner evaluation.",
8
  "version_reading_notes": [
9
  "Version 1 is the public-sample 12-task harness with minimal and neural heads.",
10
  "Version 2 is the selected 128-episode same-split simple/NN baseline alignment.",
11
- "Version 3 is the verified model-branch layer: the current final Qwen3-Omni LoRA package is the JSON-task diagnostic result, Cosmos3-Nano is a future-window compatibility result, and Cosmos3-Super Reasoner is a base-weight JSON-task evaluation; Cosmos3-Super now has a camera-pose forward-dynamics contract audit and schema-only packer smoke, but no new fine-tuned weight release."
12
  ],
13
  "versions": [
14
  {
@@ -313,16 +313,17 @@
313
  "source": "results/omni_finetune/verified_public/",
314
  "split": "episode/session held-out split; exact task target depends on backbone contract",
315
  "counts": {
316
- "verified_branch_count": 6,
317
- "qwen3_verified_package_count": 4,
318
- "cosmos3_verified_package_count": 2,
319
  "cosmos3_nano_verified_package_count": 1,
320
- "cosmos3_super_verified_package_count": 1
321
  },
322
  "models": [
323
  "Qwen3-Omni LoRA",
324
  "Cosmos3-Nano future-window compatibility branch",
325
- "Cosmos3-Super Reasoner base-weight evaluation"
 
326
  ],
327
  "branches": [
328
  {
@@ -370,6 +371,49 @@
370
  "is_current": true,
371
  "weights_repository": "planned separate Cosmos3 model repo after a real Cosmos diffusion/LoRA fine-tune exists; current result remains artifacts-only"
372
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
373
  {
374
  "id": "xperience10m_cosmos3_super_reasoner_128ep_test_full_20260607",
375
  "title": "Cosmos3-Super Reasoner",
@@ -501,6 +545,52 @@
501
  "is_current": false,
502
  "weights_repository": "historical diagnostic package; keep separate from the final 128-episode adapter repo"
503
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
504
  {
505
  "id": "xperience10m_qwen3_omni_128ep_structured_json_v2_reuse_full8gpu_lora_eval_test_full",
506
  "title": "Qwen3-Omni LoRA",
@@ -602,11 +692,75 @@
602
  "global_step": 712
603
  }
604
  ],
605
- "is_current": true,
606
- "weights_repository": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
607
  }
608
  ],
609
- "interpretation": "This layer contains the held-out foundation-model packages. Qwen3-Omni packages evaluate structured JSON task prediction; Cosmos3-Nano evaluates a future-window world-model compatibility adapter; Cosmos3-Super Reasoner evaluates staged base weights through vLLM on the JSON task. Neither Cosmos branch is a new fine-tuned weight release yet."
610
  }
611
  ],
612
  "model_groups": [
@@ -697,6 +851,146 @@
697
  "interpretation": "This validates the sensor-adapter token path on one real episode before loading or LoRA-tuning Qwen3-Omni. It is not comparable to the 128-episode held-out LoRA result."
698
  }
699
  ],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
700
  "multi_episode_128_runs": [
701
  {
702
  "id": "xperience10m_qwen3_omni_128ep_96train_16val_16test_valmon_20260605_eval",
@@ -790,6 +1084,52 @@
790
  "is_current": false,
791
  "weights_repository": "historical diagnostic package; keep separate from the final 128-episode adapter repo"
792
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
793
  {
794
  "id": "xperience10m_qwen3_omni_128ep_structured_json_v2_reuse_full8gpu_lora_eval_test_full",
795
  "title": "Qwen3-Omni LoRA",
@@ -891,11 +1231,75 @@
891
  "global_step": 712
892
  }
893
  ],
894
- "is_current": true,
895
- "weights_repository": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
896
  }
897
  ],
898
- "comparison_note": "The one-episode Qwen entry is only a sensor-adapter smoke test with Qwen3 weights unloaded. The 128-episode entries are real held-out LoRA diagnostics; the current final adapter belongs in the separate Qwen model repo."
899
  },
900
  {
901
  "id": "cosmos3_nano_world_model",
@@ -1014,6 +1418,43 @@
1014
  "weights": "none; readiness audit only, no adapter checkpoint",
1015
  "interpretation": "This probe confirms the staged Cosmos3-Super Diffusers/GPU runtime and the same JSON QA dataset are visible. It predates the camera-pose action-target export, so use the 20260608 contract audit for the current trainer-readiness status."
1016
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1017
  {
1018
  "id": "xperience10m_cosmos3_super_training_contract_audit_camera_pose_20260608",
1019
  "title": "Cosmos3-Super Camera-Pose Target Audit",
@@ -1045,7 +1486,7 @@
1045
  "weights_updated": false
1046
  },
1047
  "weights": "none; action-target contract audit only, no adapter checkpoint",
1048
- "interpretation": "The selected dataset now has valid Cosmos3 camera_pose forward_dynamics targets for an egocentric camera-motion proxy. These remove the target-schema blocker for action-conditioned world-model training, but they supervise noisy vision tokens rather than preds_action. The remaining work is a pipeline-loaded packer check and one-sample forward-dynamics overfit; action-token prediction needs a separate policy or inverse-dynamics target export."
1049
  },
1050
  {
1051
  "id": "xperience10m_cosmos3_super_action_packer_schema_smoke_20260608",
@@ -1111,18 +1552,80 @@
1111
  "weights_repository": "none for this run: staged base nv-community/Cosmos3-Super weights were evaluated through vLLM; create a separate repo only after new adapter or fine-tuned weights exist"
1112
  }
1113
  ],
1114
- "comparison_note": "Cosmos3-Super is now represented by a verified 448-window held-out Reasoner evaluation on the same JSON task as Qwen3. It uses staged base weights through vLLM, so it is a model-branch diagnostic, not a weight release. A camera-pose proxy forward-dynamics target export now passes the contract audit and schema-only packer smoke; true Cosmos3-Super fine-tuning is still not launched until the pipeline-loaded packer check and one-sample overfit exist."
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1115
  }
1116
  ],
1117
  "model_group_reading_notes": [
1118
  "Use model_groups when comparing one-episode and 128-episode artifacts within the same model family.",
1119
  "Task-head baselines have both a one-episode public-sample run and a 128-episode same-split metadata/text run.",
1120
- "Qwen3-Omni has a one-episode sensor-adapter smoke test and separate 128-episode LoRA diagnostic packages; only the final 128-episode adapter belongs in the Qwen LoRA model repo.",
1121
  "Cosmos3-Nano has a 128-episode future-window compatibility package.",
1122
- "Cosmos3-Super has a 128-episode base-weight Reasoner evaluation on the JSON task plus a camera-pose forward-dynamics contract audit; create a separate Cosmos model repo only after real Cosmos adapter/fine-tuned weights exist."
1123
  ],
1124
  "pending": [
1125
- "Use the final Qwen3 full-eval package as the current Qwen result; older Qwen package rows remain historical diagnostics for comparison.",
1126
- "Promote Cosmos3 from Nano compatibility, Super base-weight evaluation, and the camera-pose forward-dynamics contract to true fine-tuning only after the pipeline-loaded packer check and one-sample overfit produce new weights."
1127
  ]
1128
  }
 
1
  {
2
  "title": "Ropedia Xperience-10M Current Result Versions and Model Groups",
3
+ "generated_at_utc": "2026-06-11T04:42:46+00:00",
4
  "status": "pass",
5
  "version_count": 3,
6
+ "model_group_count": 5,
7
  "comparison_rule": "Compare only rows with the same scope and target. Single-episode raw-feature metrics, 128-episode metadata baselines, Qwen3 structured JSON metrics, and the two Cosmos3 targets answer different questions: Nano future-window retrieval versus Super structured JSON Reasoner evaluation.",
8
  "version_reading_notes": [
9
  "Version 1 is the public-sample 12-task harness with minimal and neural heads.",
10
  "Version 2 is the selected 128-episode same-split simple/NN baseline alignment.",
11
+ "Version 3 is the verified model-branch layer: the current final Qwen3-Omni LoRA package is the JSON-task diagnostic result, Cosmos3-Nano is a future-window compatibility result, Cosmos3-Super Reasoner is a base-weight JSON-task evaluation, and Cosmos3-Super Forward-Dynamics LoRA is the first Super fine-tuned adapter branch."
12
  ],
13
  "versions": [
14
  {
 
313
  "source": "results/omni_finetune/verified_public/",
314
  "split": "episode/session held-out split; exact task target depends on backbone contract",
315
  "counts": {
316
+ "verified_branch_count": 9,
317
+ "qwen3_verified_package_count": 6,
318
+ "cosmos3_verified_package_count": 3,
319
  "cosmos3_nano_verified_package_count": 1,
320
+ "cosmos3_super_verified_package_count": 2
321
  },
322
  "models": [
323
  "Qwen3-Omni LoRA",
324
  "Cosmos3-Nano future-window compatibility branch",
325
+ "Cosmos3-Super Reasoner base-weight evaluation",
326
+ "Cosmos3-Super forward-dynamics LoRA"
327
  ],
328
  "branches": [
329
  {
 
371
  "is_current": true,
372
  "weights_repository": "planned separate Cosmos3 model repo after a real Cosmos diffusion/LoRA fine-tune exists; current result remains artifacts-only"
373
  },
374
+ {
375
+ "id": "xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
376
+ "title": "Cosmos3-Super Forward-Dynamics LoRA",
377
+ "status": "verified",
378
+ "backbone": "cosmos3_super_forward_dynamics",
379
+ "dataset_contract": "xperience10m_camera_pose_forward_dynamics_v1",
380
+ "training_objective": "camera_pose_conditioned_future_vision_velocity_lora",
381
+ "source": "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/verified_result_summary.json",
382
+ "dataset_run_id": "xperience10m_cosmos3_camera_pose_targets_20260608",
383
+ "train_run_id": "xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608",
384
+ "eval_run_id": "xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
385
+ "counts": {
386
+ "dataset_samples": 3808,
387
+ "dataset_episodes": 119,
388
+ "split_counts": {
389
+ "test": 448,
390
+ "train": 2848,
391
+ "val": 512
392
+ },
393
+ "train_samples": 2848,
394
+ "val_samples": 512,
395
+ "eval_samples": 448,
396
+ "held_out_episode_count": 14,
397
+ "num_processes": 8
398
+ },
399
+ "primary_metrics": {
400
+ "adapter_parameter_numel": 26214400,
401
+ "held_out_episode_count": 14,
402
+ "test_forward_dynamics_mse": 3.6853174321087345,
403
+ "train_final_loss": 1.0785235166549683,
404
+ "val_forward_dynamics_mse": 4.008244896889664
405
+ },
406
+ "history": [
407
+ {
408
+ "epoch": 1,
409
+ "note": "FSDP 8-GPU LoRA over camera-pose-conditioned future vision velocity loss; adapter weights are excluded from this public package.",
410
+ "train_loss": 1.0785235166549683,
411
+ "val_loss": 4.008244896889664
412
+ }
413
+ ],
414
+ "is_current": true,
415
+ "weights_repository": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep"
416
+ },
417
  {
418
  "id": "xperience10m_cosmos3_super_reasoner_128ep_test_full_20260607",
419
  "title": "Cosmos3-Super Reasoner",
 
545
  "is_current": false,
546
  "weights_repository": "historical diagnostic package; keep separate from the final 128-episode adapter repo"
547
  },
548
+ {
549
+ "id": "xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
550
+ "title": "Qwen3-Omni LoRA",
551
+ "status": "verified",
552
+ "backbone": "qwen3_omni_lora",
553
+ "dataset_contract": "xperience10m_episode_json_qa_v1",
554
+ "training_objective": "structured_episode_understanding_json_qa",
555
+ "source": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full/verified_result_summary.json",
556
+ "dataset_run_id": "xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora",
557
+ "train_run_id": "xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora",
558
+ "eval_run_id": "xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
559
+ "counts": {
560
+ "dataset_samples": 34269,
561
+ "dataset_episodes": 119,
562
+ "split_counts": {
563
+ "test": 4032,
564
+ "train": 25629,
565
+ "val": 4608
566
+ },
567
+ "train_samples": 25629,
568
+ "val_samples": 1024,
569
+ "eval_samples": 4032,
570
+ "held_out_episode_count": 14,
571
+ "num_processes": 8
572
+ },
573
+ "primary_metrics": {
574
+ "json_validity_rate": 1.0,
575
+ "action_macro_f1": 0.002289711036077459,
576
+ "subtask_accuracy": 0.011194029850746268,
577
+ "transition_accuracy": 0.9908234126984127,
578
+ "next_action_accuracy": 0.053618594823032224,
579
+ "contact_accuracy": 0.7864583333333334,
580
+ "object_micro_f1": 0.31614599936244814,
581
+ "held_out_episode_count": 14
582
+ },
583
+ "history": [
584
+ {
585
+ "epoch": 1,
586
+ "train_loss": 0.06255286606544624,
587
+ "val_loss": 0.02668904885649681,
588
+ "global_step": 3204
589
+ }
590
+ ],
591
+ "is_current": true,
592
+ "weights_repository": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep"
593
+ },
594
  {
595
  "id": "xperience10m_qwen3_omni_128ep_structured_json_v2_reuse_full8gpu_lora_eval_test_full",
596
  "title": "Qwen3-Omni LoRA",
 
692
  "global_step": 712
693
  }
694
  ],
695
+ "is_current": false,
696
+ "weights_repository": "historical diagnostic package; keep separate from the final 128-episode adapter repo"
697
+ },
698
+ {
699
+ "id": "xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
700
+ "title": "Qwen3-Omni LoRA",
701
+ "status": "verified",
702
+ "backbone": "qwen3_omni_lora",
703
+ "dataset_contract": "xperience10m_episode_json_qa_v1",
704
+ "training_objective": "structured_episode_understanding_json_qa",
705
+ "source": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/verified_result_summary.json",
706
+ "dataset_run_id": "xperience10m_qwen3_omni_128ep_96train_16val_16test_valmon_20260605",
707
+ "train_run_id": "xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora",
708
+ "eval_run_id": "xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
709
+ "counts": {
710
+ "dataset_samples": 3808,
711
+ "dataset_episodes": 119,
712
+ "split_counts": {
713
+ "train": 2848,
714
+ "val": 512,
715
+ "test": 448
716
+ },
717
+ "train_samples": 2848,
718
+ "val_samples": 512,
719
+ "eval_samples": 448,
720
+ "held_out_episode_count": 14,
721
+ "num_processes": 8
722
+ },
723
+ "primary_metrics": {
724
+ "json_validity_rate": 1.0,
725
+ "action_macro_f1": 0.0018678269676001454,
726
+ "subtask_accuracy": 0.0,
727
+ "transition_accuracy": 0.9732142857142857,
728
+ "next_action_accuracy": 0.033482142857142856,
729
+ "contact_accuracy": 0.7299107142857143,
730
+ "object_micro_f1": 0.31099781500364165,
731
+ "held_out_episode_count": 14
732
+ },
733
+ "history": [
734
+ {
735
+ "epoch": 1,
736
+ "train_loss": 0.40796751019628613,
737
+ "val_loss": 0.03258896619081497,
738
+ "global_step": 356
739
+ },
740
+ {
741
+ "epoch": 2,
742
+ "train_loss": 0.027628723937453012,
743
+ "val_loss": 0.027754632756114006,
744
+ "global_step": 712
745
+ },
746
+ {
747
+ "epoch": 3,
748
+ "train_loss": 0.02446955946807781,
749
+ "val_loss": 0.026343274861574173,
750
+ "global_step": 1068
751
+ },
752
+ {
753
+ "epoch": 4,
754
+ "train_loss": 0.022728607045444712,
755
+ "val_loss": 0.025629229843616486,
756
+ "global_step": 1424
757
+ }
758
+ ],
759
+ "is_current": false,
760
+ "weights_repository": "historical diagnostic package; keep separate from the final 128-episode adapter repo"
761
  }
762
  ],
763
+ "interpretation": "This layer contains the held-out foundation-model packages. Qwen3-Omni packages evaluate structured JSON task prediction; Cosmos3-Nano evaluates a future-window world-model compatibility adapter; Cosmos3-Super Reasoner evaluates staged base weights through vLLM on the JSON task; Cosmos3-Super Forward-Dynamics LoRA is the first Super adapter branch and evaluates camera-pose-conditioned future vision velocity loss."
764
  }
765
  ],
766
  "model_groups": [
 
851
  "interpretation": "This validates the sensor-adapter token path on one real episode before loading or LoRA-tuning Qwen3-Omni. It is not comparable to the 128-episode held-out LoRA result."
852
  }
853
  ],
854
+ "readiness_runs": [
855
+ {
856
+ "id": "xperience10m_qwen3_omni_128ep_fullparam_smoke_preemptible_8gpu_20260609",
857
+ "title": "Full-Parameter 1-Step Feasibility Smoke",
858
+ "scope_label": "full-param gate",
859
+ "scope": "1 optimizer step over 8 train samples",
860
+ "status": "passed",
861
+ "source": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_smoke_preemptible_8gpu_20260609/fullparam_feasibility_summary.json",
862
+ "split": "selected 128-episode train split",
863
+ "counts": {
864
+ "samples": 8,
865
+ "steps": 1,
866
+ "num_processes": 8
867
+ },
868
+ "primary_metrics": {
869
+ "full_parameter_gate": "passed",
870
+ "observed_train_steps": 1,
871
+ "final_step_loss": 1.2726006507873535,
872
+ "epoch_train_loss": 1.2726006507873535,
873
+ "checkpoint_saved": false
874
+ },
875
+ "weights": "no full-parameter checkpoint or public weights; save_mode=none",
876
+ "interpretation": "Full-parameter FSDP feasibility evidence only. This gate is not a held-out model result, full fine-tune, checkpoint release, or public weight package."
877
+ },
878
+ {
879
+ "id": "xperience10m_qwen3_omni_128ep_fullparam_shorttrain8_preemptible_8gpu_20260609",
880
+ "title": "Full-Parameter 8-Step Short Train",
881
+ "scope_label": "full-param gate",
882
+ "scope": "8 optimizer steps over 64 train samples",
883
+ "status": "passed",
884
+ "source": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_shorttrain8_preemptible_8gpu_20260609/fullparam_shorttrain8_summary.json",
885
+ "split": "selected 128-episode train split",
886
+ "counts": {
887
+ "samples": 64,
888
+ "steps": 8,
889
+ "num_processes": 8
890
+ },
891
+ "primary_metrics": {
892
+ "full_parameter_gate": "passed",
893
+ "observed_train_steps": 8,
894
+ "final_step_loss": 1.180522084236145,
895
+ "epoch_train_loss": 1.2190196067094803,
896
+ "checkpoint_saved": false
897
+ },
898
+ "weights": "no full-parameter checkpoint or public weights; save_mode=none",
899
+ "interpretation": "Full-parameter FSDP feasibility evidence only. This gate is not a held-out model result, full fine-tune, checkpoint release, or public weight package."
900
+ },
901
+ {
902
+ "id": "xperience10m_qwen3_omni_128ep_fullparam_pilot32_preemptible_8gpu_20260609",
903
+ "title": "Full-Parameter 32-Step Pilot",
904
+ "scope_label": "full-param gate",
905
+ "scope": "32 optimizer steps over 256 train samples",
906
+ "status": "passed",
907
+ "source": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot32_preemptible_8gpu_20260609/fullparam_pilot32_summary.json",
908
+ "split": "selected 128-episode train split",
909
+ "counts": {
910
+ "samples": 256,
911
+ "steps": 32,
912
+ "num_processes": 8
913
+ },
914
+ "primary_metrics": {
915
+ "full_parameter_gate": "passed",
916
+ "observed_train_steps": 32,
917
+ "final_step_loss": 0.2206273376941681,
918
+ "epoch_train_loss": 0.8451133379712701,
919
+ "checkpoint_saved": false
920
+ },
921
+ "weights": "no full-parameter checkpoint or public weights; save_mode=none",
922
+ "interpretation": "Full-parameter FSDP feasibility evidence only. This gate is not a held-out model result, full fine-tune, checkpoint release, or public weight package."
923
+ },
924
+ {
925
+ "id": "xperience10m_qwen3_omni_128ep_fullparam_pilot64_preemptible_8gpu_20260609",
926
+ "title": "Full-Parameter 64-Step Pilot",
927
+ "scope_label": "full-param gate",
928
+ "scope": "64 optimizer steps over 512 train samples",
929
+ "status": "passed",
930
+ "source": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot64_preemptible_8gpu_20260609/fullparam_pilot64_summary.json",
931
+ "split": "selected 128-episode train split",
932
+ "counts": {
933
+ "samples": 512,
934
+ "steps": 64,
935
+ "num_processes": 8
936
+ },
937
+ "primary_metrics": {
938
+ "full_parameter_gate": "passed",
939
+ "observed_train_steps": 64,
940
+ "final_step_loss": 0.011219973675906658,
941
+ "epoch_train_loss": 0.4434075650788145,
942
+ "checkpoint_saved": false
943
+ },
944
+ "weights": "no full-parameter checkpoint or public weights; save_mode=none",
945
+ "interpretation": "Full-parameter FSDP feasibility evidence only. This gate is not a held-out model result, full fine-tune, checkpoint release, or public weight package."
946
+ },
947
+ {
948
+ "id": "xperience10m_qwen3_omni_128ep_fullparam_pilot128_preemptible_8gpu_20260609",
949
+ "title": "Full-Parameter 128-Step Opportunistic Pilot",
950
+ "scope_label": "full-param gate",
951
+ "scope": "planned 128 optimizer steps over 1024 train samples; preempted for Qwen v5 handoff",
952
+ "status": "preempted_for_qwen_v5_handoff",
953
+ "source": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot128_preemptible_8gpu_20260609/fullparam_pilot128_summary.json",
954
+ "split": "selected 128-episode train split",
955
+ "counts": {
956
+ "samples": 1024,
957
+ "steps": 0,
958
+ "num_processes": 8
959
+ },
960
+ "primary_metrics": {
961
+ "full_parameter_gate": "preempted_for_qwen_v5_handoff",
962
+ "observed_train_steps": 0,
963
+ "final_step_loss": null,
964
+ "epoch_train_loss": null,
965
+ "checkpoint_saved": false
966
+ },
967
+ "weights": "no full-parameter checkpoint or public weights; save_mode=none",
968
+ "interpretation": "Full-parameter FSDP feasibility evidence only. This gate is not a held-out model result, full fine-tune, checkpoint release, or public weight package."
969
+ },
970
+ {
971
+ "id": "xperience10m_qwen3_omni_128ep_fullparam_pilot128_after_qwen_v5_preemptible_8gpu_20260609",
972
+ "title": "Full-Parameter 128-Step Post-Qwen-v5 Pilot",
973
+ "scope_label": "full-param gate",
974
+ "scope": "128 optimizer steps over 1024 train samples after verified Qwen v5 handoff",
975
+ "status": "passed",
976
+ "source": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot128_after_qwen_v5_preemptible_8gpu_20260609/training_metadata.json",
977
+ "split": "selected 128-episode train split",
978
+ "counts": {
979
+ "samples": 1024,
980
+ "steps": 128,
981
+ "num_processes": 8
982
+ },
983
+ "primary_metrics": {
984
+ "full_parameter_gate": "passed",
985
+ "observed_train_steps": 128,
986
+ "final_step_loss": 0.0136940386146307,
987
+ "epoch_train_loss": 0.21579630990163423,
988
+ "checkpoint_saved": false
989
+ },
990
+ "weights": "no full-parameter checkpoint or public weights; save_mode=none",
991
+ "interpretation": "Full-parameter FSDP feasibility evidence only. This gate is not a held-out model result, full fine-tune, checkpoint release, or public weight package."
992
+ }
993
+ ],
994
  "multi_episode_128_runs": [
995
  {
996
  "id": "xperience10m_qwen3_omni_128ep_96train_16val_16test_valmon_20260605_eval",
 
1084
  "is_current": false,
1085
  "weights_repository": "historical diagnostic package; keep separate from the final 128-episode adapter repo"
1086
  },
1087
+ {
1088
+ "id": "xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
1089
+ "title": "Qwen3-Omni LoRA",
1090
+ "status": "verified",
1091
+ "backbone": "qwen3_omni_lora",
1092
+ "dataset_contract": "xperience10m_episode_json_qa_v1",
1093
+ "training_objective": "structured_episode_understanding_json_qa",
1094
+ "source": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full/verified_result_summary.json",
1095
+ "dataset_run_id": "xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora",
1096
+ "train_run_id": "xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora",
1097
+ "eval_run_id": "xperience10m_qwen3_omni_128ep_multiscale_cap96_v5_full8gpu_lora_eval_test_full",
1098
+ "counts": {
1099
+ "dataset_samples": 34269,
1100
+ "dataset_episodes": 119,
1101
+ "split_counts": {
1102
+ "test": 4032,
1103
+ "train": 25629,
1104
+ "val": 4608
1105
+ },
1106
+ "train_samples": 25629,
1107
+ "val_samples": 1024,
1108
+ "eval_samples": 4032,
1109
+ "held_out_episode_count": 14,
1110
+ "num_processes": 8
1111
+ },
1112
+ "primary_metrics": {
1113
+ "json_validity_rate": 1.0,
1114
+ "action_macro_f1": 0.002289711036077459,
1115
+ "subtask_accuracy": 0.011194029850746268,
1116
+ "transition_accuracy": 0.9908234126984127,
1117
+ "next_action_accuracy": 0.053618594823032224,
1118
+ "contact_accuracy": 0.7864583333333334,
1119
+ "object_micro_f1": 0.31614599936244814,
1120
+ "held_out_episode_count": 14
1121
+ },
1122
+ "history": [
1123
+ {
1124
+ "epoch": 1,
1125
+ "train_loss": 0.06255286606544624,
1126
+ "val_loss": 0.02668904885649681,
1127
+ "global_step": 3204
1128
+ }
1129
+ ],
1130
+ "is_current": true,
1131
+ "weights_repository": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep"
1132
+ },
1133
  {
1134
  "id": "xperience10m_qwen3_omni_128ep_structured_json_v2_reuse_full8gpu_lora_eval_test_full",
1135
  "title": "Qwen3-Omni LoRA",
 
1231
  "global_step": 712
1232
  }
1233
  ],
1234
+ "is_current": false,
1235
+ "weights_repository": "historical diagnostic package; keep separate from the final 128-episode adapter repo"
1236
+ },
1237
+ {
1238
+ "id": "xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1239
+ "title": "Qwen3-Omni LoRA",
1240
+ "status": "verified",
1241
+ "backbone": "qwen3_omni_lora",
1242
+ "dataset_contract": "xperience10m_episode_json_qa_v1",
1243
+ "training_objective": "structured_episode_understanding_json_qa",
1244
+ "source": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/verified_result_summary.json",
1245
+ "dataset_run_id": "xperience10m_qwen3_omni_128ep_96train_16val_16test_valmon_20260605",
1246
+ "train_run_id": "xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora",
1247
+ "eval_run_id": "xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
1248
+ "counts": {
1249
+ "dataset_samples": 3808,
1250
+ "dataset_episodes": 119,
1251
+ "split_counts": {
1252
+ "train": 2848,
1253
+ "val": 512,
1254
+ "test": 448
1255
+ },
1256
+ "train_samples": 2848,
1257
+ "val_samples": 512,
1258
+ "eval_samples": 448,
1259
+ "held_out_episode_count": 14,
1260
+ "num_processes": 8
1261
+ },
1262
+ "primary_metrics": {
1263
+ "json_validity_rate": 1.0,
1264
+ "action_macro_f1": 0.0018678269676001454,
1265
+ "subtask_accuracy": 0.0,
1266
+ "transition_accuracy": 0.9732142857142857,
1267
+ "next_action_accuracy": 0.033482142857142856,
1268
+ "contact_accuracy": 0.7299107142857143,
1269
+ "object_micro_f1": 0.31099781500364165,
1270
+ "held_out_episode_count": 14
1271
+ },
1272
+ "history": [
1273
+ {
1274
+ "epoch": 1,
1275
+ "train_loss": 0.40796751019628613,
1276
+ "val_loss": 0.03258896619081497,
1277
+ "global_step": 356
1278
+ },
1279
+ {
1280
+ "epoch": 2,
1281
+ "train_loss": 0.027628723937453012,
1282
+ "val_loss": 0.027754632756114006,
1283
+ "global_step": 712
1284
+ },
1285
+ {
1286
+ "epoch": 3,
1287
+ "train_loss": 0.02446955946807781,
1288
+ "val_loss": 0.026343274861574173,
1289
+ "global_step": 1068
1290
+ },
1291
+ {
1292
+ "epoch": 4,
1293
+ "train_loss": 0.022728607045444712,
1294
+ "val_loss": 0.025629229843616486,
1295
+ "global_step": 1424
1296
+ }
1297
+ ],
1298
+ "is_current": false,
1299
+ "weights_repository": "historical diagnostic package; keep separate from the final 128-episode adapter repo"
1300
  }
1301
  ],
1302
+ "comparison_note": "The one-episode Qwen entry is only a sensor-adapter smoke test with Qwen3 weights unloaded. The 128-episode entries are real held-out LoRA diagnostics; the current final adapter belongs in the separate Qwen model repo. The full-parameter rows are feasibility gates only and intentionally publish no checkpoints or full-parameter weights."
1303
  },
1304
  {
1305
  "id": "cosmos3_nano_world_model",
 
1418
  "weights": "none; readiness audit only, no adapter checkpoint",
1419
  "interpretation": "This probe confirms the staged Cosmos3-Super Diffusers/GPU runtime and the same JSON QA dataset are visible. It predates the camera-pose action-target export, so use the 20260608 contract audit for the current trainer-readiness status."
1420
  },
1421
+ {
1422
+ "id": "xperience10m_cosmos3_super_training_readiness_metadata_a100_20260609",
1423
+ "title": "Cosmos3-Super Remote Staging Readiness Probe",
1424
+ "scope_label": "staging readiness",
1425
+ "scope": "secondary 4-GPU staging tree, JSON-task dataset visibility, and metadata-only Cosmos3-Super runtime probe",
1426
+ "status": "blocked_until_trainer_implemented",
1427
+ "source": "results/omni_finetune/xperience10m_cosmos3_super_training_readiness_metadata_a100_20260609/training_readiness.json",
1428
+ "split": "train/val/test by selected episode/session",
1429
+ "counts": {
1430
+ "dataset_samples": 3808,
1431
+ "split_counts": {
1432
+ "test": {
1433
+ "samples": 448,
1434
+ "episodes": 14,
1435
+ "actions": 189
1436
+ },
1437
+ "train": {
1438
+ "samples": 2848,
1439
+ "episodes": 89,
1440
+ "actions": 885
1441
+ },
1442
+ "val": {
1443
+ "samples": 512,
1444
+ "episodes": 16,
1445
+ "actions": 223
1446
+ }
1447
+ }
1448
+ },
1449
+ "primary_metrics": {
1450
+ "model_files_visible": false,
1451
+ "diffusers_runtime_supported": false,
1452
+ "cuda_device_count": 4,
1453
+ "weights_updated": false
1454
+ },
1455
+ "weights": "none; staging readiness audit only, no adapter checkpoint",
1456
+ "interpretation": "This metadata-only probe checks the secondary 4-GPU staging tree without loading the model pipeline or updating weights. It confirms the JSON task dataset is present, but the Cosmos3-Super model files and Diffusers runtime are not staged there yet, so real Super training should wait for model/runtime staging or run on the already prepared main host."
1457
+ },
1458
  {
1459
  "id": "xperience10m_cosmos3_super_training_contract_audit_camera_pose_20260608",
1460
  "title": "Cosmos3-Super Camera-Pose Target Audit",
 
1486
  "weights_updated": false
1487
  },
1488
  "weights": "none; action-target contract audit only, no adapter checkpoint",
1489
+ "interpretation": "The selected dataset now has valid Cosmos3 camera_pose forward_dynamics targets for an egocentric camera-motion proxy. These remove the target-schema blocker for action-conditioned world-model training, but they supervise noisy vision tokens rather than preds_action. The remaining work is a trainable Cosmos3-Super implementation that can backpropagate through this loss surface at the required memory scale; action-token prediction needs a separate policy or inverse-dynamics target export."
1490
  },
1491
  {
1492
  "id": "xperience10m_cosmos3_super_action_packer_schema_smoke_20260608",
 
1552
  "weights_repository": "none for this run: staged base nv-community/Cosmos3-Super weights were evaluated through vLLM; create a separate repo only after new adapter or fine-tuned weights exist"
1553
  }
1554
  ],
1555
+ "comparison_note": "Cosmos3-Super is now represented by a verified 448-window held-out Reasoner evaluation on the same JSON task as Qwen3. It uses staged base weights through vLLM, so it is a model-branch diagnostic, not a weight release. A camera-pose proxy forward-dynamics target export now passes the contract audit and schema-only packer smoke; the separate Forward-Dynamics LoRA group records the trainable adapter run and loss-based held-out evaluation."
1556
+ },
1557
+ {
1558
+ "id": "cosmos3_super_forward_dynamics",
1559
+ "model_family": "Cosmos3-Super Forward-Dynamics LoRA",
1560
+ "model_type": "PEFT LoRA over nv-community/Cosmos3-Super for camera-pose-conditioned future vision velocity",
1561
+ "weight_repository": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep",
1562
+ "one_episode_runs": [
1563
+ {
1564
+ "id": "cosmos3_super_forward_dynamics_overfit_smoke",
1565
+ "title": "Cosmos3-Super Forward-Dynamics Overfit Smoke",
1566
+ "scope": "small overfit smoke before 128-episode scale-up",
1567
+ "status": "verified_smoke",
1568
+ "source": "results/omni_finetune/xperience10m_cosmos3_super_forward_dynamics_lora_overfit_after_qwen_v4_20260608_fsdp8_attn256_gradfix_savefix2/",
1569
+ "weights": "local repaired LoRA smoke adapter, not public packaged as final",
1570
+ "interpretation": "Validated the trainable adapter path, FSDP save repair, and Diffusers load before the full 128-episode run."
1571
+ }
1572
+ ],
1573
+ "multi_episode_128_runs": [
1574
+ {
1575
+ "id": "xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1576
+ "title": "Cosmos3-Super Forward-Dynamics LoRA",
1577
+ "status": "verified",
1578
+ "backbone": "cosmos3_super_forward_dynamics",
1579
+ "dataset_contract": "xperience10m_camera_pose_forward_dynamics_v1",
1580
+ "training_objective": "camera_pose_conditioned_future_vision_velocity_lora",
1581
+ "source": "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/verified_result_summary.json",
1582
+ "dataset_run_id": "xperience10m_cosmos3_camera_pose_targets_20260608",
1583
+ "train_run_id": "xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608",
1584
+ "eval_run_id": "xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp",
1585
+ "counts": {
1586
+ "dataset_samples": 3808,
1587
+ "dataset_episodes": 119,
1588
+ "split_counts": {
1589
+ "test": 448,
1590
+ "train": 2848,
1591
+ "val": 512
1592
+ },
1593
+ "train_samples": 2848,
1594
+ "val_samples": 512,
1595
+ "eval_samples": 448,
1596
+ "held_out_episode_count": 14,
1597
+ "num_processes": 8
1598
+ },
1599
+ "primary_metrics": {
1600
+ "adapter_parameter_numel": 26214400,
1601
+ "held_out_episode_count": 14,
1602
+ "test_forward_dynamics_mse": 3.6853174321087345,
1603
+ "train_final_loss": 1.0785235166549683,
1604
+ "val_forward_dynamics_mse": 4.008244896889664
1605
+ },
1606
+ "history": [
1607
+ {
1608
+ "epoch": 1,
1609
+ "note": "FSDP 8-GPU LoRA over camera-pose-conditioned future vision velocity loss; adapter weights are excluded from this public package.",
1610
+ "train_loss": 1.0785235166549683,
1611
+ "val_loss": 4.008244896889664
1612
+ }
1613
+ ],
1614
+ "is_current": true,
1615
+ "weights_repository": "https://huggingface.co/cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep"
1616
+ }
1617
+ ],
1618
+ "comparison_note": "This is the first verified Cosmos3-Super fine-tuned adapter branch. Its metric is forward-dynamics MSE, so compare it to world-model loss or future-prediction targets, not to Qwen JSON classification accuracy."
1619
  }
1620
  ],
1621
  "model_group_reading_notes": [
1622
  "Use model_groups when comparing one-episode and 128-episode artifacts within the same model family.",
1623
  "Task-head baselines have both a one-episode public-sample run and a 128-episode same-split metadata/text run.",
1624
+ "Qwen3-Omni has a one-episode sensor-adapter smoke test, full-parameter feasibility gates, and separate 128-episode LoRA diagnostic packages; the newest verified full-eval 128-episode adapter belongs in the Qwen LoRA model repo.",
1625
  "Cosmos3-Nano has a 128-episode future-window compatibility package.",
1626
+ "Cosmos3-Super now has both a 128-episode base-weight Reasoner evaluation on the JSON task and a fine-tuned forward-dynamics LoRA branch over camera-pose proxy targets."
1627
  ],
1628
  "pending": [
1629
+ "Use the verified Qwen3 v5 dense multiscale full-eval package as the current Qwen row; older Qwen package rows remain historical diagnostics for comparison."
 
1630
  ]
1631
  }
data/project_packet.json CHANGED
@@ -1,26 +1,28 @@
1
  {
2
- "title": "Ropedia Xperience-10M Task Suite Project Packet",
3
- "version": "2026-06-01",
4
- "scope_status": {
5
- "validated_data": "one public Xperience-10M sample episode",
6
- "aligned_frames": 5821,
7
- "sliding_windows": 1161,
8
- "current_feature_dimensions": 8546,
9
- "core_task_count": 12,
10
- "neural_head_count": 12,
11
- "direction_extension_probe_count": 4,
12
- "raw_xperience10m_data_in_repo": false,
13
- "audio_feature_status": "Audio is one of the synchronized source modalities in the current task representation.",
14
- "qwen3_omni_32_episode_claim": false,
15
- "qwen3_omni_status": "The selected 96/16/16 Qwen3-Omni final diagnostic result is verified, meets the strict-JSON target, and still has weak action/subtask metrics that guide the next error-analysis pass."
16
- },
17
- "reading_path": [
18
- {
19
- "step": 1,
20
- "question": "What is the current project scope?",
21
- "primary_artifacts": [
22
- "PROJECT_STATUS.md",
23
- "docs/data/project_status.json",
 
 
24
  "RESEARCH_ROADMAP.md",
25
  "docs/data/research_roadmap.json",
26
  "EVIDENCE_CONTRACT.md",
@@ -36,24 +38,24 @@
36
  "docs/data/figure_index.json",
37
  "docs/data/source_alignment_audit.json",
38
  "docs/data/xperience10m_dataset_card_alignment.json",
39
- "docs/data/mirror_parity.json",
40
- "docs/data/publication_audit.json",
41
- "docs/data/scope_claims_audit.json",
42
- "docs/data/website_integrity.json"
43
- ],
44
- "readout": "The project status table and roadmap give the compact current-state summary. Single-episode task engineering, metrics, visualizations, public website integrity, mirror parity, same-split 128-episode baselines, the final selected-episode Qwen3-Omni diagnostic result, the Cosmos3-Nano compatibility package, the Cosmos3-Super base-weight Reasoner evaluation, and the Cosmos3-Super camera-pose forward-dynamics contract audit plus schema-only packer smoke are implemented; stronger action/subtask and real Cosmos fine-tuned model quality remain follow-ups."
45
- },
46
- {
47
- "step": 2,
48
  "question": "What do the official Xperience-10M dataset and sample cards say?",
49
- "primary_artifacts": [
50
- "XPERIENCE10M_DATASET_CARD_ALIGNMENT.md",
51
- "docs/data/xperience10m_dataset_card_alignment.json",
52
  "https://huggingface.co/datasets/ropedia-ai/xperience-10m",
53
  "https://huggingface.co/datasets/ropedia-ai/xperience-10m-sample"
54
  ],
55
  "readout": "The full upstream dataset is a manually gated large-scale 4D multimodal egocentric source. The public sample card records the sample license, HOMIE Toolkit path, and Rerun 0.29.0 visualization path. This repo validates one public sample episode and lists the current project coverage."
56
- },
57
  {
58
  "step": 3,
59
  "question": "Are source facts consistently presented?",
@@ -77,48 +79,59 @@
77
  {
78
  "step": 5,
79
  "question": "How can the public pipeline be reproduced?",
80
- "primary_artifacts": [
81
- "REPRODUCIBILITY.md",
82
- "docs/data/reproducibility_matrix.json",
83
- "notes/reproducibility_audit.md"
84
- ],
85
- "readout": "The public sample pipeline has explicit commands, expected outputs, and a prior exact-match reproduction check over the committed metrics."
86
- },
87
  {
88
  "step": 6,
89
  "question": "What is inside one model input?",
90
- "primary_artifacts": [
91
- "results/episode_task_suite/windows.csv",
92
- "results/episode_task_suite/feature_manifest.json",
93
- "results/episode_task_suite/available_modalities.json",
94
- "docs/data/modality_atlas.json"
95
- ],
96
- "readout": "The current model input is an 8,546-dimensional aligned multimodal window, and the readable atlas shows each public-sample modality without raw data redistribution."
97
- },
98
  {
99
  "step": 7,
100
  "question": "Do the task metrics have committed evidence?",
101
- "primary_artifacts": [
102
- "results/episode_task_suite/summary_report.json",
103
- "results/episode_task_suite/neural_mlp/",
104
- "docs/data/summary_metrics.json"
105
- ],
106
- "readout": "Each of the 12 tasks has minimal-head metrics and a matching neural MLP result over the same window contracts."
107
- },
108
- {
109
- "step": 8,
110
  "question": "What is the scale-up path?",
111
- "primary_artifacts": [
112
- "RESEARCH_ROADMAP.md",
113
- "docs/data/research_roadmap.json",
114
- "results/omni_finetune/DATA_ACCESS_STATUS.md",
115
- "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md",
116
- "scripts/omni/discover_xperience10m_sources.py",
117
- "docs/data/omni_finetune_verified_result.json"
118
- ],
119
- "readout": "The selected-episode held-out Qwen3-Omni final diagnostic result is verified and JSON-format reliability meets the 98% target. The next milestone is action/subtask error analysis and a stronger model-quality run on the same split."
120
- }
121
- ],
 
 
 
 
 
 
 
 
 
 
 
122
  "project_status": "PROJECT_STATUS.md",
123
  "project_status_json": "docs/data/project_status.json",
124
  "research_roadmap": "RESEARCH_ROADMAP.md",
@@ -128,23 +141,27 @@
128
  "source_alignment_audit": "SOURCE_ALIGNMENT_AUDIT.md",
129
  "source_alignment_audit_json": "docs/data/source_alignment_audit.json",
130
  "artifact_guide": "ARTIFACT_GUIDE.md",
131
- "artifact_index": "docs/data/artifact_index.json",
132
- "brand_assets": "docs/data/brand_assets.json",
133
- "figure_index": "FIGURE_INDEX.md",
134
- "figure_index_json": "docs/data/figure_index.json",
135
- "reproducibility_matrix": "docs/data/reproducibility_matrix.json",
136
- "public_surfaces": {
137
- "github_repo": "https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite",
138
- "github_pages": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/",
139
- "hf_space": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite",
140
- "hf_static_app": "https://cy0307-ropedia-xperience-10m-task-suite.static.hf.space/",
141
- "hf_artifacts": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts",
142
- "hf_model_baselines": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines"
143
- },
144
- "current_reading_notes": [
145
- "The first cross-episode Qwen3-Omni diagnostic pilot is verified, but strong model quality is not yet shown.",
146
- "Older Qwen3-Omni setup artifacts are separate from the verified selected-episode diagnostic package.",
147
- "Feature-vector reconstruction is separate from pixel-depth, mesh, NeRF, or Gaussian reconstruction.",
148
- "Raw Xperience-10M data is not redistributed in this repo."
149
- ]
 
 
 
 
150
  }
 
1
  {
2
+ "title": "Ropedia Xperience-10M Task Suite Project Packet",
3
+ "version": "2026-06-08",
4
+ "scope_status": {
5
+ "validated_data": "one public Xperience-10M sample episode",
6
+ "aligned_frames": 5821,
7
+ "sliding_windows": 1161,
8
+ "current_feature_dimensions": 8546,
9
+ "core_task_count": 12,
10
+ "neural_head_count": 12,
11
+ "direction_extension_probe_count": 4,
12
+ "raw_xperience10m_data_in_repo": false,
13
+ "audio_feature_status": "Audio is one of the synchronized source modalities in the current task representation.",
14
+ "qwen3_omni_32_episode_claim": false,
15
+ "qwen3_omni_status": "The selected 96/16/16 Qwen3-Omni v4 final diagnostic result is verified, meets the strict-JSON target, and still has weak action/subtask metrics that guide the next error-analysis pass.",
16
+ "cosmos3_super_forward_dynamics_lora_status": "The first Cosmos3-Super fine-tuned adapter branch is verified as a forward-dynamics LoRA over camera-pose proxy targets; it reports loss metrics, not JSON action-label accuracy.",
17
+ "task_suite_enhancement_128_status": "Current no-new-episode enhancement pack recommends multiscale_20s10_40s20_80s40, hierarchical action/subtask targets, label-normalized scoring, and raw-feature shards before adding more episodes."
18
+ },
19
+ "reading_path": [
20
+ {
21
+ "step": 1,
22
+ "question": "What is the current project scope?",
23
+ "primary_artifacts": [
24
+ "PROJECT_STATUS.md",
25
+ "docs/data/project_status.json",
26
  "RESEARCH_ROADMAP.md",
27
  "docs/data/research_roadmap.json",
28
  "EVIDENCE_CONTRACT.md",
 
38
  "docs/data/figure_index.json",
39
  "docs/data/source_alignment_audit.json",
40
  "docs/data/xperience10m_dataset_card_alignment.json",
41
+ "docs/data/mirror_parity.json",
42
+ "docs/data/publication_audit.json",
43
+ "docs/data/scope_claims_audit.json",
44
+ "docs/data/website_integrity.json"
45
+ ],
46
+ "readout": "The project status table and roadmap give the compact current-state summary. Single-episode task engineering, metrics, visualizations, public website integrity, mirror parity, same-split 128-episode baselines, the final selected-episode Qwen3-Omni diagnostic result, the Cosmos3-Nano compatibility package, the Cosmos3-Super base-weight Reasoner evaluation, and the Cosmos3-Super Forward-Dynamics LoRA package are implemented; stronger action/subtask quality and policy-compatible action targets remain follow-ups."
47
+ },
48
+ {
49
+ "step": 2,
50
  "question": "What do the official Xperience-10M dataset and sample cards say?",
51
+ "primary_artifacts": [
52
+ "XPERIENCE10M_DATASET_CARD_ALIGNMENT.md",
53
+ "docs/data/xperience10m_dataset_card_alignment.json",
54
  "https://huggingface.co/datasets/ropedia-ai/xperience-10m",
55
  "https://huggingface.co/datasets/ropedia-ai/xperience-10m-sample"
56
  ],
57
  "readout": "The full upstream dataset is a manually gated large-scale 4D multimodal egocentric source. The public sample card records the sample license, HOMIE Toolkit path, and Rerun 0.29.0 visualization path. This repo validates one public sample episode and lists the current project coverage."
58
+ },
59
  {
60
  "step": 3,
61
  "question": "Are source facts consistently presented?",
 
79
  {
80
  "step": 5,
81
  "question": "How can the public pipeline be reproduced?",
82
+ "primary_artifacts": [
83
+ "REPRODUCIBILITY.md",
84
+ "docs/data/reproducibility_matrix.json",
85
+ "notes/reproducibility_audit.md"
86
+ ],
87
+ "readout": "The public sample pipeline has explicit commands, expected outputs, and a prior exact-match reproduction check over the committed metrics."
88
+ },
89
  {
90
  "step": 6,
91
  "question": "What is inside one model input?",
92
+ "primary_artifacts": [
93
+ "results/episode_task_suite/windows.csv",
94
+ "results/episode_task_suite/feature_manifest.json",
95
+ "results/episode_task_suite/available_modalities.json",
96
+ "docs/data/modality_atlas.json"
97
+ ],
98
+ "readout": "The current model input is an 8,546-dimensional aligned multimodal window, and the readable atlas shows each public-sample modality without raw data redistribution."
99
+ },
100
  {
101
  "step": 7,
102
  "question": "Do the task metrics have committed evidence?",
103
+ "primary_artifacts": [
104
+ "results/episode_task_suite/summary_report.json",
105
+ "results/episode_task_suite/neural_mlp/",
106
+ "docs/data/summary_metrics.json"
107
+ ],
108
+ "readout": "Each of the 12 tasks has minimal-head metrics and a matching neural MLP result over the same window contracts."
109
+ },
110
+ {
111
+ "step": 8,
112
  "question": "What is the scale-up path?",
113
+ "primary_artifacts": [
114
+ "RESEARCH_ROADMAP.md",
115
+ "docs/data/research_roadmap.json",
116
+ "results/omni_finetune/DATA_ACCESS_STATUS.md",
117
+ "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md",
118
+ "scripts/omni/discover_xperience10m_sources.py",
119
+ "docs/data/omni_finetune_verified_result.json"
120
+ ],
121
+ "readout": "The selected-episode held-out Qwen3-Omni final diagnostic result is verified and JSON-format reliability meets the 98% target. The same public comparison also includes the verified 128-episode baselines, Cosmos3-Nano compatibility result, Cosmos3-Super Reasoner evaluation, and Cosmos3-Super Forward-Dynamics LoRA package. The next milestone is action/subtask error analysis and stronger model-quality runs on the same split."
122
+ },
123
+ {
124
+ "step": 9,
125
+ "question": "How can the current 128 episodes be pushed harder?",
126
+ "primary_artifacts": [
127
+ "TASK_SUITE_ENHANCEMENT_128.md",
128
+ "docs/data/task_suite_enhancement_128.json",
129
+ "results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json",
130
+ "results/omni_finetune/task_suite_enhancement_128_v1_20260608/dense_window_scenarios.csv"
131
+ ],
132
+ "readout": "The current selected split can be expanded with dense and multiscale windows without adding episodes. The recommended export target is multiscale_20s10_40s20_80s40, followed by hierarchical action/subtask targets and raw-feature shards for unsupported tasks."
133
+ }
134
+ ],
135
  "project_status": "PROJECT_STATUS.md",
136
  "project_status_json": "docs/data/project_status.json",
137
  "research_roadmap": "RESEARCH_ROADMAP.md",
 
141
  "source_alignment_audit": "SOURCE_ALIGNMENT_AUDIT.md",
142
  "source_alignment_audit_json": "docs/data/source_alignment_audit.json",
143
  "artifact_guide": "ARTIFACT_GUIDE.md",
144
+ "artifact_index": "docs/data/artifact_index.json",
145
+ "brand_assets": "docs/data/brand_assets.json",
146
+ "figure_index": "FIGURE_INDEX.md",
147
+ "figure_index_json": "docs/data/figure_index.json",
148
+ "reproducibility_matrix": "docs/data/reproducibility_matrix.json",
149
+ "public_surfaces": {
150
+ "github_repo": "https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite",
151
+ "github_pages": "https://chaoyue0307.github.io/ropedia-xperience-10m-task-suite/",
152
+ "hf_space": "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite",
153
+ "hf_static_app": "https://cy0307-ropedia-xperience-10m-task-suite.static.hf.space/",
154
+ "hf_artifacts": "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts",
155
+ "hf_model_baselines": "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines"
156
+ },
157
+ "current_reading_notes": [
158
+ "The first cross-episode Qwen3-Omni v4 diagnostic pilot is verified, but strong model quality is not yet shown; action/subtask metrics remain weak.",
159
+ "The current 128-episode suite has a no-new-episode enhancement plan: multiscale_20s10_40s20_80s40 windows, hierarchical labels, label-normalized scoring, and raw-feature shard export.",
160
+ "Cosmos3-Super Forward-Dynamics LoRA is verified as a loss-based world-model adapter branch, not as JSON action-token prediction.",
161
+ "Older Qwen3-Omni setup artifacts are separate from the verified selected-episode diagnostic package.",
162
+ "Feature-vector reconstruction is separate from pixel-depth, mesh, NeRF, or Gaussian reconstruction.",
163
+ "Raw Xperience-10M data is not redistributed in this repo."
164
+ ],
165
+ "task_suite_enhancement_128": "TASK_SUITE_ENHANCEMENT_128.md",
166
+ "task_suite_enhancement_128_json": "docs/data/task_suite_enhancement_128.json"
167
  }
data/project_status.json CHANGED
@@ -1,299 +1,339 @@
1
  {
2
- "title": "Ropedia Xperience-10M Task Suite Project Status",
3
- "version": "2026-06-01",
4
- "decision": "public_sample_pipeline_verified_128_aligned_baselines_qwen3_cosmos_comparison",
5
- "research_positioning": "A research-engineering study that makes one public Xperience-10M sample episode inspectable, defines embodied-AI tasks over synchronized modalities, records baseline behavior, aligns simple/NN baselines to the selected 128-episode split, and compares verified Qwen3-Omni and Cosmos3 branch packages as early cross-episode diagnostics.",
6
- "scope_boundary": {
7
- "validated_episode_count": 1,
8
- "aligned_frames": 5821,
9
- "sliding_windows": 1161,
10
- "current_feature_dimensions": 8546,
11
- "core_task_count": 12,
12
- "neural_head_count": 12,
13
- "direction_extension_probe_count": 4,
14
- "audio_featurized": true,
15
- "raw_xperience10m_data_redistributed": false,
16
- "qwen3_omni_32_episode_claim": false,
17
- "qwen3_omni_verified_diagnostic_pilot": true,
18
- "qwen3_omni_selected_episode_counts": {
19
- "train": 96,
20
- "val": 16,
21
- "test": 16
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
22
  },
23
- "qwen3_omni_exported_window_counts": {
24
- "train": 2848,
25
- "val": 512,
26
- "test": 448
27
- },
28
- "qwen3_omni_json_validity_rate": 0.9977678571428571,
29
- "qwen3_omni_validation_aware": true,
30
- "qwen3_omni_json_quality_target_met": true,
31
- "qwen3_omni_lora_adapter_repo": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep",
32
- "cosmos3_nano_future_window_compatibility_verified": true,
33
- "cosmos3_nano_future_window_test_predictions": 378,
34
- "cosmos3_super_reasoner_verified": true,
35
- "cosmos3_super_reasoner_test_predictions": 448,
36
- "cosmos3_super_reasoner_json_validity_rate": 0.5111607142857143,
37
- "omni_model_comparison_available": true,
38
- "multi_episode_128_aligned_baselines": true,
39
- "multi_episode_128_baseline_window_counts": {
40
- "train": 2848,
41
- "val": 512,
42
- "test": 448
43
- },
44
- "multi_episode_128_baseline_task_count": 12
45
- },
46
- "rows": [
47
- {
48
- "area": "Public-sample pipeline",
49
- "status": "verified",
50
- "evidence": [
51
- "results/episode_task_suite/summary_report.json",
52
- "results/episode_task_suite/windows.csv",
53
- "results/episode_task_suite/feature_manifest.json"
54
- ],
55
- "readout": "One public Xperience-10M sample episode is converted into 5,821 frames, 1,161 aligned 20-frame windows, and an 8,546-dimensional representation for repeatable task evaluation."
56
- },
57
- {
58
- "area": "Task suite",
59
- "status": "verified",
60
- "evidence": [
61
- "scripts/episode_task_suite.py",
62
- "results/episode_task_suite/",
63
- "docs/data/summary_metrics.json"
64
- ],
65
- "readout": "All 12 task contracts have committed metrics, predictions, and minimal baseline outputs."
66
- },
67
- {
68
- "area": "Neural heads",
69
- "status": "verified",
70
- "evidence": [
71
- "scripts/neural_task_models.py",
72
- "results/episode_task_suite/neural_mlp/"
73
- ],
74
- "readout": "Each task also has a compact PyTorch MLP run over the same feature tensor and chronological split."
75
- },
76
- {
77
- "area": "Audio contribution study",
78
- "status": "verified",
79
- "evidence": [
80
- "scripts/audio_ablation_and_raw_upgrade.py",
81
- "results/audio_ablation/",
82
- "docs/data/audio_ablation_summary.json"
83
- ],
84
- "readout": "Audio variants improve the primary metric on 6 of 12 task contracts in this single-episode setting."
85
- },
86
- {
87
- "area": "Evaluation protocol",
88
- "status": "verified",
89
- "evidence": [
90
- "EVALUATION_PROTOCOL.md",
91
- "docs/data/evaluation_protocol.json",
92
- "scripts/build_evaluation_protocol.py"
93
- ],
94
- "readout": "Windowing, chronological split, per-task metrics, leakage controls, and current limitations are generated from committed metric artifacts."
95
- },
96
- {
97
- "area": "Research takeaways",
98
- "status": "verified",
99
- "evidence": [
100
- "RESEARCH_TAKEAWAYS.md",
101
- "docs/data/research_takeaways.json",
102
- "scripts/build_research_takeaways.py"
103
- ],
104
- "readout": "The main result interpretation is generated from committed metrics: chronological class shift, neural gains on dynamics/order/alignment, open retrieval/reconstruction problems, and the need for held-out episodes."
105
- },
106
- {
107
- "area": "Research roadmap",
108
- "status": "current",
109
- "evidence": [
110
- "RESEARCH_ROADMAP.md",
111
- "docs/data/research_roadmap.json"
112
- ],
113
- "readout": "The roadmap connects public-sample task development to the final verified Qwen3-Omni diagnostic result, same-split baseline alignment, action/subtask error analysis, robustness runs, world/policy branches, and the future Xperience-native pretraining goal."
114
- },
115
- {
116
- "area": "Foundation-model plan",
117
- "status": "current",
118
- "evidence": [
119
- "FOUNDATION_MODEL_PLAN.md",
120
- "docs/data/foundation_model_plan.json"
121
- ],
122
- "readout": "Qwen3-Omni remains the first trainable held-out LoRA baseline; Cosmos 3 is now represented by a verified Cosmos3-Nano future-window compatibility package, a verified Cosmos3-Super base-weight Reasoner evaluation, and a Cosmos3-Super camera-pose proxy forward-dynamics contract audit plus schema-only packer smoke. The current target supports vision-velocity training under action conditioning, not supervised action-token prediction; OpenVLA/openpi/GR00T are policy candidates after robot-compatible action targets are explicit."
123
- },
124
- {
125
- "area": "Omni model extension contract",
126
- "status": "current",
127
- "evidence": [
128
- "OMNI_MODEL_EXTENSION_CONTRACT.md",
129
- "configs/omni_backbones/",
130
- "scripts/omni/backbone_registry.py",
131
- "scripts/omni/smoke_test_backbone_packaging.py"
132
- ],
133
- "readout": "Future Qwen, Cosmos-style, and VLA/policy branches must keep the same episode split discipline, held-out metrics, validation gate, public-safe package contract, and explicit forbidden-artifact policy before reporting results."
134
- },
135
- {
136
- "area": "Xperience Embodied Foundation Model",
137
- "status": "future_goal",
138
- "evidence": [
139
- "XPERIENCE_EMBODIED_FOUNDATION_MODEL_PRETRAINING.md"
140
- ],
141
- "readout": "A future full-corpus pretraining plan describes target modules, objectives, staged scale-up, hardware ranges, and evaluation for a domain-specific embodied foundation model."
142
- },
143
- {
144
- "area": "Official dataset wording",
145
- "status": "verified",
146
- "evidence": [
147
- "XPERIENCE10M_DATASET_CARD_ALIGNMENT.md",
148
- "docs/data/xperience10m_dataset_card_alignment.json"
149
- ],
150
- "readout": "Public wording is aligned to the official gated Xperience-10M dataset card, public sample card, and HF API metadata, including modalities, scale, access path, sample license/tooling, and current project coverage."
151
- },
152
- {
153
- "area": "Source alignment",
154
- "status": "verified",
155
- "evidence": [
156
- "SOURCE_ALIGNMENT_AUDIT.md",
157
- "docs/data/source_alignment_audit.json",
158
- "scripts/validate_source_alignment.py"
159
- ],
160
- "readout": "Source facts, sample details, API-listing notes, and project coverage are checked across repo docs, website, and HF cards."
161
- },
162
- {
163
- "area": "Website and HF mirrors",
164
- "status": "verified",
165
- "evidence": [
166
- "docs/data/website_integrity.json",
167
- "docs/data/mirror_parity.json",
168
- "docs/data/live_publication_status.json"
169
- ],
170
- "readout": "Local website links/assets pass, prepared mirrors match, and public GitHub/HF URLs have been checked after upload."
171
- },
172
- {
173
- "area": "Publication package",
174
- "status": "verified",
175
- "evidence": [
176
- "docs/data/publication_audit.json",
177
- "QUALITY_GATES.md",
178
- "docs/data/quality_gates.json"
179
- ],
180
- "readout": "Public bundles are checked for raw-data exclusion, cache exclusion, heavy-archive exclusion, credential-text checks, and current presentation assets."
181
- },
182
- {
183
- "area": "Reproducibility",
184
- "status": "verified_for_public_sample",
185
- "evidence": [
186
- "REPRODUCIBILITY.md",
187
- "docs/data/reproducibility_matrix.json",
188
- "notes/reproducibility_audit.md"
189
- ],
190
- "readout": "The public sample workflow has explicit commands, expected outputs, and exact-match reproduction evidence."
191
- },
192
- {
193
- "area": "128-episode aligned baselines",
194
- "status": "verified_companion_result",
195
- "evidence": [
196
- "results/omni_finetune/multi_episode_128_task_baselines/BASELINE_ALIGNMENT_REPORT.md",
197
- "results/omni_finetune/multi_episode_128_task_baselines/summary_report.json",
198
- "scripts/omni/run_128_task_baselines.py"
199
- ],
200
- "readout": "The earlier simple and neural baseline framing is aligned to the selected 96/16/16 episode split used by the Qwen3-Omni pilot. JSON-supported tasks have metadata/text simple and neural MLP metrics; raw-feature-only tasks are explicitly marked unsupported until 128-run sensor feature blocks are available."
201
- },
202
- {
203
- "area": "Current result comparison",
204
- "status": "verified_generated_summary",
205
- "evidence": [
206
- "docs/data/omni_model_comparison.json",
207
- "results/omni_finetune/OMNI_MODEL_COMPARISON.md",
208
- "scripts/omni/build_omni_model_comparison.py"
209
- ],
210
- "readout": "The public comparison now has two views: the three result layers and a model-family grouping. The model grouping pairs 1-episode and 128-episode entries for task-head baselines, separates Qwen3-Omni sensor-adapter smoke from 128-episode LoRA diagnostics, and separates Cosmos3-Nano future-window compatibility from Cosmos3-Super base-weight Reasoner evaluation."
211
- },
212
- {
213
- "area": "Qwen3-Omni fine-tuning",
214
- "status": "final_verified_diagnostic_result_json_target_met",
215
- "evidence": [
216
- "docs/data/omni_finetune_verified_result.json",
217
- "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v3_strict_label_prompt_reuse_lora_eval_test_full/",
218
- "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep",
219
- "scripts/omni/package_verified_omni_result.py",
220
- "scripts/omni/audit_verified_omni_package.py",
221
- "scripts/omni/analyze_qwen3_omni_errors.py"
222
- ],
223
- "readout": "The selected 96/16/16 episode split now has a v3 strict-label public-safe held-out package with 3,808 exported windows, 512 validation windows, 448 test predictions, two training epochs reused from the same LoRA adapter, validation/audit summaries, and a public LoRA adapter repo. JSON validity is 100.00%, meeting the 98% target; transition accuracy is 97.32%, contact accuracy is 72.10%, object micro-F1 is 30.69%, and action/subtask metrics remain weak, so it is still a diagnostic baseline rather than a strong model-quality claim."
224
- },
225
- {
226
- "area": "Cosmos3-Nano future-window branch",
227
- "status": "verified_compatibility_result",
228
- "evidence": [
229
- "configs/omni_backbones/cosmos_world_model.json",
230
- "scripts/omni/export_cosmos3_future_window_dataset.py",
231
- "scripts/omni/eval_cosmos3_future_window_retrieval.py",
232
- "results/omni_finetune/verified_public/xperience10m_cosmos3_nano_128ep_future_window_h5_compat_adapter_eval_test_full/verified_result_summary.json"
233
- ],
234
- "readout": "The Cosmos3-Nano branch now has a public-safe verified future-window compatibility package with 3,213 future-window samples, 378 held-out test predictions, future retrieval MRR 0.0221, temporal consistency 0.0952, transition accuracy 0.9683, and contact accuracy 0.7434. It is a compatibility adapter result, not a full Cosmos diffusion-weight fine-tune."
235
- },
236
- {
237
- "area": "Cosmos3-Super Reasoner branch",
238
- "status": "verified_base_weight_result",
239
- "evidence": [
240
- "configs/omni_backbones/cosmos3_super_reasoner.json",
241
- "scripts/omni/eval_cosmos3_super_reasoner.py",
242
- "scripts/omni/run_cosmos3_super_reasoner_eval.sh",
243
- "results/omni_finetune/verified_public/xperience10m_cosmos3_super_reasoner_128ep_test_full_20260607/verified_result_summary.json"
244
- ],
245
- "readout": "Cosmos3-Super Reasoner now has a public-safe verified 448-window held-out evaluation on the same structured JSON task as Qwen3. It uses staged nv-community/Cosmos3-Super base weights through an 8-GPU vLLM server, not fine-tuned weights: JSON validity 0.5112, action macro-F1 0.0008, transition accuracy 0.3683, contact accuracy 0.3214, and object micro-F1 0.1370."
246
- },
247
- {
248
- "area": "Cosmos3-Super action-target contract",
249
- "status": "ready_for_forward_dynamics_trainer_implementation",
250
- "evidence": [
251
- "scripts/omni/export_cosmos3_camera_pose_targets.py",
252
- "scripts/omni/pack_cosmos3_super_action_batch.py",
253
- "results/omni_finetune/xperience10m_cosmos3_camera_pose_targets_20260608/target_manifest.json",
254
- "results/omni_finetune/xperience10m_cosmos3_super_training_contract_audit_camera_pose_20260608/training_contract_audit.json",
255
- "results/omni_finetune/xperience10m_cosmos3_super_action_packer_schema_smoke_20260608/packer_summary.json"
256
- ],
257
- "readout": "The selected 128-episode JSONL is augmented with 3,808/3,808 valid camera_pose proxy cosmos_action_target records from SLAM pose deltas. The schema-only packer smoke confirms the current forward_dynamics target should supervise noisy vision tokens under camera-pose conditioning; it does not supervise preds_action. Remaining work is a pipeline-loaded packer check, one-sample forward-dynamics overfit, and a separate policy/inverse target export before claiming action-token prediction."
258
- },
259
- {
260
- "area": "Raw Xperience-10M redistribution",
261
- "status": "not_included",
262
- "evidence": [
263
- "DATA_NOTICE.md",
264
- "docs/data/publication_audit.json"
265
- ],
266
- "readout": "Raw MP4, HDF5, RRD files, private gated data, and full Qwen weights are intentionally excluded."
267
- }
268
- ],
269
- "fast_research_route": [
270
- "Read PROJECT_STATUS.md and EVIDENCE_CONTRACT.md to establish what is implemented.",
271
- "Open docs/data/project_packet.json for the machine-readable project path.",
272
- "Inspect RESEARCH_TAKEAWAYS.md and docs/data/research_takeaways.json before interpreting model scores.",
273
- "Inspect RESEARCH_ROADMAP.md and docs/data/research_roadmap.json for the path from public-sample task work to multi-episode modeling.",
274
- "Inspect FOUNDATION_MODEL_PLAN.md and docs/data/foundation_model_plan.json before choosing a backbone branch.",
275
- "Inspect OMNI_MODEL_EXTENSION_CONTRACT.md and run python scripts/omni/backbone_registry.py --validate --json before adding a new Qwen, Cosmos-style, or VLA/policy branch.",
276
- "Inspect XPERIENCE_EMBODIED_FOUNDATION_MODEL_PRETRAINING.md for the long-term full-corpus pretraining goal.",
277
- "Inspect docs/data/summary_metrics.json and results/episode_task_suite/neural_mlp/ to check the 12-task outputs.",
278
- "Inspect results/audio_ablation/AUDIO_ABLATION_SUMMARY.md before judging whether audio helps the current task suite.",
279
- "Inspect EVALUATION_PROTOCOL.md before judging task metrics or leakage controls.",
280
- "Inspect SOURCE_ALIGNMENT_AUDIT.md before judging source-card consistency across public surfaces.",
281
- "Inspect XPERIENCE10M_DATASET_CARD_ALIGNMENT.md before judging dataset wording.",
282
- "Inspect results/omni_finetune/multi_episode_128_task_baselines/BASELINE_ALIGNMENT_REPORT.md before comparing simple/NN baselines to the selected 128-episode setup.",
283
- "Inspect docs/data/omni_model_comparison.json before comparing the current three result versions or the model-family 1-episode versus 128-episode groupings.",
284
- "Inspect docs/data/omni_finetune_verified_result.json before judging the Qwen3-Omni diagnostic pilot."
285
- ],
286
- "current_reading_notes": [
287
- "The final Qwen3-Omni diagnostic result is verified and meets the strict-JSON target, but action/subtask held-out quality is still weak.",
288
- "Use docs/data/omni_model_comparison.json to compare both views: the single-episode/128-baseline/model-branch result layers and the model-family grouping for task heads, Qwen3-Omni LoRA, Cosmos3-Nano, and Cosmos3-Super.",
289
- "Use docs/data/omni_finetune_verified_result.json and the latest verified_public final Qwen package for current held-out results.",
290
- "The 128-episode aligned simple/NN baselines use metadata/text features from the derived Qwen JSONL export; they align the split and task ids but do not replace raw-modality baselines for trajectory, retrieval, reconstruction, or misalignment tasks.",
291
- "The Cosmos3-Nano future-window branch is verified as a compatibility adapter result, Cosmos3-Super Reasoner is verified as a base-weight evaluation, and Cosmos3-Super camera-pose forward-dynamics targets now pass the contract audit plus a schema-only packer smoke; one-episode Cosmos fine-tuning and full Cosmos adapter/diffusion-weight fine-tuning remain pending, so no Cosmos weight repo should be published yet.",
292
- "The current reconstruction task reconstructs feature vectors, not pixel-depth, mesh, NeRF, or Gaussian reconstruction.",
293
- "Audio is one of the synchronized source modalities in the current task representation.",
294
- "The audio ablation report compares audio/no-audio variants across all 12 task contracts in results/audio_ablation/.",
295
- "Foundation-model selection is explicit: Qwen3-Omni is the immediate trainable pilot, Cosmos 3 is the first world-model branch, Cosmos3-Super has a camera-pose proxy forward-dynamics contract ready for trainer implementation, and policy models such as OpenVLA/openpi/GR00T wait for robot-compatible action-target conversion.",
296
- "Future model branches should be added through the backbone registry and verified package contract, not as one-off result folders with incompatible metrics or publication rules.",
297
- "The Xperience Embodied Foundation Model is a future native-pretraining goal, not a completed model or current benchmark."
298
- ]
 
299
  }
 
1
  {
2
+ "title": "Ropedia Xperience-10M Task Suite Project Status",
3
+ "version": "2026-06-08",
4
+ "decision": "public_sample_pipeline_verified_128_enhancement_qwen3_v4_cosmos_comparison",
5
+ "research_positioning": "A research-engineering study that makes one public Xperience-10M sample episode inspectable, defines embodied-AI tasks over synchronized modalities, records baseline behavior, aligns simple/NN baselines to the selected 128-episode split, compares verified Qwen3-Omni and Cosmos3 branch packages as early cross-episode diagnostics, and now records a no-new-episode enhancement pack for pushing the current 128-episode suite harder.",
6
+ "scope_boundary": {
7
+ "validated_episode_count": 1,
8
+ "aligned_frames": 5821,
9
+ "sliding_windows": 1161,
10
+ "current_feature_dimensions": 8546,
11
+ "core_task_count": 12,
12
+ "neural_head_count": 12,
13
+ "direction_extension_probe_count": 4,
14
+ "audio_featurized": true,
15
+ "raw_xperience10m_data_redistributed": false,
16
+ "qwen3_omni_32_episode_claim": false,
17
+ "qwen3_omni_verified_diagnostic_pilot": true,
18
+ "qwen3_omni_selected_episode_counts": {
19
+ "train": 96,
20
+ "val": 16,
21
+ "test": 16
22
+ },
23
+ "qwen3_omni_exported_window_counts": {
24
+ "train": 2848,
25
+ "val": 512,
26
+ "test": 448
27
+ },
28
+ "qwen3_omni_json_validity_rate": 1.0,
29
+ "qwen3_omni_validation_aware": true,
30
+ "qwen3_omni_json_quality_target_met": true,
31
+ "qwen3_omni_lora_adapter_repo": "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep",
32
+ "cosmos3_nano_future_window_compatibility_verified": true,
33
+ "cosmos3_nano_future_window_test_predictions": 378,
34
+ "cosmos3_super_reasoner_verified": true,
35
+ "cosmos3_super_reasoner_test_predictions": 448,
36
+ "cosmos3_super_reasoner_json_validity_rate": 0.5111607142857143,
37
+ "cosmos3_super_forward_dynamics_lora_verified": true,
38
+ "cosmos3_super_forward_dynamics_train_rows": 2848,
39
+ "cosmos3_super_forward_dynamics_val_rows": 512,
40
+ "cosmos3_super_forward_dynamics_test_rows": 448,
41
+ "cosmos3_super_forward_dynamics_test_mse": 3.6853174321087345,
42
+ "cosmos3_super_forward_dynamics_adapter_params": 26214400,
43
+ "omni_model_comparison_available": true,
44
+ "multi_episode_128_aligned_baselines": true,
45
+ "multi_episode_128_baseline_window_counts": {
46
+ "train": 2848,
47
+ "val": 512,
48
+ "test": 448
49
+ },
50
+ "multi_episode_128_baseline_task_count": 12,
51
+ "qwen3_omni_current_eval_run_id": "xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
52
+ "qwen3_omni_current_train_epochs": 4,
53
+ "qwen3_omni_action_macro_f1": 0.0018678269676001454,
54
+ "qwen3_omni_subtask_accuracy": 0.0,
55
+ "qwen3_omni_contact_accuracy": 0.7299107142857143,
56
+ "qwen3_omni_object_micro_f1": 0.31099781500364165,
57
+ "task_suite_enhancement_128_available": true,
58
+ "task_suite_enhancement_128_current_windows": 3808,
59
+ "task_suite_enhancement_128_recommended_export": "multiscale_20s10_40s20_80s40",
60
+ "task_suite_enhancement_128_estimated_windows": 106095
61
  },
62
+ "rows": [
63
+ {
64
+ "area": "Public-sample pipeline",
65
+ "status": "verified",
66
+ "evidence": [
67
+ "results/episode_task_suite/summary_report.json",
68
+ "results/episode_task_suite/windows.csv",
69
+ "results/episode_task_suite/feature_manifest.json"
70
+ ],
71
+ "readout": "One public Xperience-10M sample episode is converted into 5,821 frames, 1,161 aligned 20-frame windows, and an 8,546-dimensional representation for repeatable task evaluation."
72
+ },
73
+ {
74
+ "area": "Task suite",
75
+ "status": "verified",
76
+ "evidence": [
77
+ "scripts/episode_task_suite.py",
78
+ "results/episode_task_suite/",
79
+ "docs/data/summary_metrics.json"
80
+ ],
81
+ "readout": "All 12 task contracts have committed metrics, predictions, and minimal baseline outputs."
82
+ },
83
+ {
84
+ "area": "Neural heads",
85
+ "status": "verified",
86
+ "evidence": [
87
+ "scripts/neural_task_models.py",
88
+ "results/episode_task_suite/neural_mlp/"
89
+ ],
90
+ "readout": "Each task also has a compact PyTorch MLP run over the same feature tensor and chronological split."
91
+ },
92
+ {
93
+ "area": "Audio contribution study",
94
+ "status": "verified",
95
+ "evidence": [
96
+ "scripts/audio_ablation_and_raw_upgrade.py",
97
+ "results/audio_ablation/",
98
+ "docs/data/audio_ablation_summary.json"
99
+ ],
100
+ "readout": "Audio variants improve the primary metric on 6 of 12 task contracts in this single-episode setting."
101
+ },
102
+ {
103
+ "area": "Evaluation protocol",
104
+ "status": "verified",
105
+ "evidence": [
106
+ "EVALUATION_PROTOCOL.md",
107
+ "docs/data/evaluation_protocol.json",
108
+ "scripts/build_evaluation_protocol.py"
109
+ ],
110
+ "readout": "Windowing, chronological split, per-task metrics, leakage controls, and current limitations are generated from committed metric artifacts."
111
+ },
112
+ {
113
+ "area": "Research takeaways",
114
+ "status": "verified",
115
+ "evidence": [
116
+ "RESEARCH_TAKEAWAYS.md",
117
+ "docs/data/research_takeaways.json",
118
+ "scripts/build_research_takeaways.py"
119
+ ],
120
+ "readout": "The main result interpretation is generated from committed metrics: chronological class shift, neural gains on dynamics/order/alignment, open retrieval/reconstruction problems, and the need for held-out episodes."
121
+ },
122
+ {
123
+ "area": "Research roadmap",
124
+ "status": "current",
125
+ "evidence": [
126
+ "RESEARCH_ROADMAP.md",
127
+ "docs/data/research_roadmap.json"
128
+ ],
129
+ "readout": "The roadmap connects public-sample task development to the final verified Qwen3-Omni diagnostic result, same-split baseline alignment, the no-new-episode 128-suite enhancement pack, action/subtask error analysis, robustness runs, world/policy branches, and the future Xperience-native pretraining goal."
130
+ },
131
+ {
132
+ "area": "128-episode task-suite enhancement pack",
133
+ "status": "current_no_new_episode_plan",
134
+ "evidence": [
135
+ "TASK_SUITE_ENHANCEMENT_128.md",
136
+ "docs/data/task_suite_enhancement_128.json",
137
+ "results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json",
138
+ "scripts/omni/build_task_suite_enhancement_128.py"
139
+ ],
140
+ "readout": "The current 3,808-window selected split can be stressed without more episodes by exporting denser and multiscale windows. The recommended next export is multiscale_20s10_40s20_80s40, estimated at 106,095 windows from observed frame spans; the pack also defines hierarchical action/subtask targets, raw-feature shard priorities for unsupported tasks, and Qwen/Cosmos follow-up run cards."
141
+ },
142
+ {
143
+ "area": "Foundation-model plan",
144
+ "status": "current",
145
+ "evidence": [
146
+ "FOUNDATION_MODEL_PLAN.md",
147
+ "docs/data/foundation_model_plan.json"
148
+ ],
149
+ "readout": "Qwen3-Omni remains the first structured JSON LoRA baseline; Cosmos 3 is now represented by a verified Cosmos3-Nano future-window compatibility package, a verified Cosmos3-Super base-weight Reasoner evaluation, and a verified Cosmos3-Super Forward-Dynamics LoRA over camera-pose proxy targets. The Super LoRA target supports vision-velocity training under action conditioning, not supervised action-token prediction; OpenVLA/openpi/GR00T remain policy candidates after robot-compatible action targets are explicit."
150
+ },
151
+ {
152
+ "area": "Omni model extension contract",
153
+ "status": "current",
154
+ "evidence": [
155
+ "OMNI_MODEL_EXTENSION_CONTRACT.md",
156
+ "configs/omni_backbones/",
157
+ "scripts/omni/backbone_registry.py",
158
+ "scripts/omni/smoke_test_backbone_packaging.py"
159
+ ],
160
+ "readout": "Future Qwen, Cosmos-style, and VLA/policy branches must keep the same episode split discipline, held-out metrics, validation gate, public-safe package contract, and explicit forbidden-artifact policy before reporting results."
161
+ },
162
+ {
163
+ "area": "Xperience Embodied Foundation Model",
164
+ "status": "future_goal",
165
+ "evidence": [
166
+ "XPERIENCE_EMBODIED_FOUNDATION_MODEL_PRETRAINING.md"
167
+ ],
168
+ "readout": "A future full-corpus pretraining plan describes target modules, objectives, staged scale-up, hardware ranges, and evaluation for a domain-specific embodied foundation model."
169
+ },
170
+ {
171
+ "area": "Official dataset wording",
172
+ "status": "verified",
173
+ "evidence": [
174
+ "XPERIENCE10M_DATASET_CARD_ALIGNMENT.md",
175
+ "docs/data/xperience10m_dataset_card_alignment.json"
176
+ ],
177
+ "readout": "Public wording is aligned to the official gated Xperience-10M dataset card, public sample card, and HF API metadata, including modalities, scale, access path, sample license/tooling, and current project coverage."
178
+ },
179
+ {
180
+ "area": "Source alignment",
181
+ "status": "verified",
182
+ "evidence": [
183
+ "SOURCE_ALIGNMENT_AUDIT.md",
184
+ "docs/data/source_alignment_audit.json",
185
+ "scripts/validate_source_alignment.py"
186
+ ],
187
+ "readout": "Source facts, sample details, API-listing notes, and project coverage are checked across repo docs, website, and HF cards."
188
+ },
189
+ {
190
+ "area": "Website and HF mirrors",
191
+ "status": "verified",
192
+ "evidence": [
193
+ "docs/data/website_integrity.json",
194
+ "docs/data/mirror_parity.json",
195
+ "docs/data/live_publication_status.json"
196
+ ],
197
+ "readout": "Local website links/assets pass, prepared mirrors match, and public GitHub/HF URLs have been checked after upload."
198
+ },
199
+ {
200
+ "area": "Publication package",
201
+ "status": "verified",
202
+ "evidence": [
203
+ "docs/data/publication_audit.json",
204
+ "QUALITY_GATES.md",
205
+ "docs/data/quality_gates.json"
206
+ ],
207
+ "readout": "Public bundles are checked for raw-data exclusion, cache exclusion, heavy-archive exclusion, credential-text checks, and current presentation assets."
208
+ },
209
+ {
210
+ "area": "Reproducibility",
211
+ "status": "verified_for_public_sample",
212
+ "evidence": [
213
+ "REPRODUCIBILITY.md",
214
+ "docs/data/reproducibility_matrix.json",
215
+ "notes/reproducibility_audit.md"
216
+ ],
217
+ "readout": "The public sample workflow has explicit commands, expected outputs, and exact-match reproduction evidence."
218
+ },
219
+ {
220
+ "area": "128-episode aligned baselines",
221
+ "status": "verified_companion_result",
222
+ "evidence": [
223
+ "results/omni_finetune/multi_episode_128_task_baselines/BASELINE_ALIGNMENT_REPORT.md",
224
+ "results/omni_finetune/multi_episode_128_task_baselines/summary_report.json",
225
+ "scripts/omni/run_128_task_baselines.py"
226
+ ],
227
+ "readout": "The earlier simple and neural baseline framing is aligned to the selected 96/16/16 episode split used by the Qwen3-Omni pilot. JSON-supported tasks have metadata/text simple and neural MLP metrics; raw-feature-only tasks are explicitly marked unsupported until 128-run sensor feature blocks are available."
228
+ },
229
+ {
230
+ "area": "Current result comparison",
231
+ "status": "verified_generated_summary",
232
+ "evidence": [
233
+ "docs/data/omni_model_comparison.json",
234
+ "results/omni_finetune/OMNI_MODEL_COMPARISON.md",
235
+ "scripts/omni/build_omni_model_comparison.py"
236
+ ],
237
+ "readout": "The public comparison now has two views: the three result layers and a model-family grouping. The model grouping pairs 1-episode and 128-episode entries for task-head baselines, separates Qwen3-Omni sensor-adapter smoke from 128-episode LoRA diagnostics, separates Cosmos3-Nano future-window compatibility from Cosmos3-Super base-weight Reasoner evaluation, and adds Cosmos3-Super Forward-Dynamics LoRA as a loss-based fine-tuned adapter branch."
238
+ },
239
+ {
240
+ "area": "Qwen3-Omni fine-tuning",
241
+ "status": "final_verified_diagnostic_result_json_target_met",
242
+ "evidence": [
243
+ "docs/data/omni_finetune_verified_result.json",
244
+ "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/",
245
+ "https://huggingface.co/cy0307/ropedia-qwen3-omni-lora-128ep",
246
+ "scripts/omni/package_verified_omni_result.py",
247
+ "scripts/omni/audit_verified_omni_package.py",
248
+ "scripts/omni/analyze_qwen3_omni_errors.py"
249
+ ],
250
+ "readout": "The selected 96/16/16 episode split now has a v4 four-epoch public-safe held-out package with 3,808 exported windows, 512 validation windows, 448 test predictions, validation/audit summaries, and a public LoRA adapter repo. JSON validity is 100.00%, meeting the 98% target; transition accuracy is 97.32%, contact accuracy is 72.99%, object micro-F1 is 31.10%, next-action accuracy is 3.35%, and action/subtask metrics remain weak, so it is still a diagnostic baseline rather than a strong model-quality claim."
251
+ },
252
+ {
253
+ "area": "Cosmos3-Nano future-window branch",
254
+ "status": "verified_compatibility_result",
255
+ "evidence": [
256
+ "configs/omni_backbones/cosmos_world_model.json",
257
+ "scripts/omni/export_cosmos3_future_window_dataset.py",
258
+ "scripts/omni/eval_cosmos3_future_window_retrieval.py",
259
+ "results/omni_finetune/verified_public/xperience10m_cosmos3_nano_128ep_future_window_h5_compat_adapter_eval_test_full/verified_result_summary.json"
260
+ ],
261
+ "readout": "The Cosmos3-Nano branch now has a public-safe verified future-window compatibility package with 3,213 future-window samples, 378 held-out test predictions, future retrieval MRR 0.0221, temporal consistency 0.0952, transition accuracy 0.9683, and contact accuracy 0.7434. It is a compatibility adapter result, not a full Cosmos diffusion-weight fine-tune."
262
+ },
263
+ {
264
+ "area": "Cosmos3-Super Reasoner branch",
265
+ "status": "verified_base_weight_result",
266
+ "evidence": [
267
+ "configs/omni_backbones/cosmos3_super_reasoner.json",
268
+ "scripts/omni/eval_cosmos3_super_reasoner.py",
269
+ "scripts/omni/run_cosmos3_super_reasoner_eval.sh",
270
+ "results/omni_finetune/verified_public/xperience10m_cosmos3_super_reasoner_128ep_test_full_20260607/verified_result_summary.json"
271
+ ],
272
+ "readout": "Cosmos3-Super Reasoner now has a public-safe verified 448-window held-out evaluation on the same structured JSON task as Qwen3. It uses staged nv-community/Cosmos3-Super base weights through an 8-GPU vLLM server, not fine-tuned weights: JSON validity 0.5112, action macro-F1 0.0008, transition accuracy 0.3683, contact accuracy 0.3214, and object micro-F1 0.1370."
273
+ },
274
+ {
275
+ "area": "Cosmos3-Super action-target contract",
276
+ "status": "superseded_by_verified_forward_dynamics_lora",
277
+ "evidence": [
278
+ "scripts/omni/export_cosmos3_camera_pose_targets.py",
279
+ "scripts/omni/pack_cosmos3_super_action_batch.py",
280
+ "results/omni_finetune/xperience10m_cosmos3_camera_pose_targets_20260608/target_manifest.json",
281
+ "results/omni_finetune/xperience10m_cosmos3_super_training_contract_audit_camera_pose_20260608/training_contract_audit.json",
282
+ "results/omni_finetune/xperience10m_cosmos3_super_action_packer_schema_smoke_20260608/packer_summary.json"
283
+ ],
284
+ "readout": "The selected 128-episode JSONL is augmented with 3,808/3,808 valid camera_pose proxy cosmos_action_target records from SLAM pose deltas. The contract and packer smoke enabled the verified forward-dynamics LoRA run; it supervises noisy vision tokens under camera-pose conditioning and does not supervise preds_action."
285
+ },
286
+ {
287
+ "area": "Cosmos3-Super Forward-Dynamics LoRA",
288
+ "status": "verified_fine_tuned_adapter_result",
289
+ "evidence": [
290
+ "configs/omni_backbones/cosmos3_super_forward_dynamics.json",
291
+ "scripts/omni/train_cosmos3_super_forward_dynamics_lora.py",
292
+ "scripts/omni/eval_cosmos3_super_forward_dynamics_lora.py",
293
+ "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/verified_result_summary.json",
294
+ "results/omni_finetune/verified_public/xperience10m_cosmos3_super_forward_dynamics_lora_128ep_train1epoch_256_attn_full8gpu_20260608_eval_test_full_fsdp/package_audit.json"
295
+ ],
296
+ "readout": "The first fine-tuned Cosmos3-Super adapter branch is verified as a public-safe package: 8-GPU FSDP LoRA, 26.2M adapter parameters, 2,848 train rows, 512 validation rows, 448 held-out test rows, validation MSE 4.0082, and test MSE 3.6853. The package excludes adapter safetensors; weights are published separately at cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep."
297
+ },
298
+ {
299
+ "area": "Raw Xperience-10M redistribution",
300
+ "status": "not_included",
301
+ "evidence": [
302
+ "DATA_NOTICE.md",
303
+ "docs/data/publication_audit.json"
304
+ ],
305
+ "readout": "Raw MP4, HDF5, RRD files, private gated data, and full Qwen weights are intentionally excluded."
306
+ }
307
+ ],
308
+ "fast_research_route": [
309
+ "Read PROJECT_STATUS.md and EVIDENCE_CONTRACT.md to establish what is implemented.",
310
+ "Open docs/data/project_packet.json for the machine-readable project path.",
311
+ "Inspect RESEARCH_TAKEAWAYS.md and docs/data/research_takeaways.json before interpreting model scores.",
312
+ "Inspect RESEARCH_ROADMAP.md and docs/data/research_roadmap.json for the path from public-sample task work to multi-episode modeling.",
313
+ "Inspect FOUNDATION_MODEL_PLAN.md and docs/data/foundation_model_plan.json before choosing a backbone branch.",
314
+ "Inspect OMNI_MODEL_EXTENSION_CONTRACT.md and run python scripts/omni/backbone_registry.py --validate --json before adding a new Qwen, Cosmos-style, or VLA/policy branch.",
315
+ "Inspect XPERIENCE_EMBODIED_FOUNDATION_MODEL_PRETRAINING.md for the long-term full-corpus pretraining goal.",
316
+ "Inspect docs/data/summary_metrics.json and results/episode_task_suite/neural_mlp/ to check the 12-task outputs.",
317
+ "Inspect results/audio_ablation/AUDIO_ABLATION_SUMMARY.md before judging whether audio helps the current task suite.",
318
+ "Inspect EVALUATION_PROTOCOL.md before judging task metrics or leakage controls.",
319
+ "Inspect SOURCE_ALIGNMENT_AUDIT.md before judging source-card consistency across public surfaces.",
320
+ "Inspect XPERIENCE10M_DATASET_CARD_ALIGNMENT.md before judging dataset wording.",
321
+ "Inspect results/omni_finetune/multi_episode_128_task_baselines/BASELINE_ALIGNMENT_REPORT.md before comparing simple/NN baselines to the selected 128-episode setup.",
322
+ "Inspect TASK_SUITE_ENHANCEMENT_128.md and docs/data/task_suite_enhancement_128.json before deciding whether more episodes are needed; the current recommended no-new-episode export is multiscale_20s10_40s20_80s40.",
323
+ "Inspect docs/data/omni_model_comparison.json before comparing the current three result versions or the model-family 1-episode versus 128-episode groupings.",
324
+ "Inspect docs/data/omni_finetune_verified_result.json before judging the Qwen3-Omni diagnostic pilot."
325
+ ],
326
+ "current_reading_notes": [
327
+ "The final Qwen3-Omni v4 diagnostic result is verified and meets the strict-JSON target, but action/subtask held-out quality is still weak: JSON validity is 100.00%, action macro-F1 is 0.0019, and subtask accuracy is 0.0000.",
328
+ "Use TASK_SUITE_ENHANCEMENT_128.md and docs/data/task_suite_enhancement_128.json to push the current 128-episode suite without more raw episodes through multiscale_20s10_40s20_80s40, hierarchical labels, label-normalized scoring, and raw-feature shard export.",
329
+ "Use docs/data/omni_model_comparison.json to compare both views: the single-episode/128-baseline/model-branch result layers and the model-family grouping for task heads, Qwen3-Omni LoRA, Cosmos3-Nano, and Cosmos3-Super.",
330
+ "The 128-episode aligned simple/NN baselines use metadata/text features from the derived Qwen JSONL export; they align the split and task ids but do not replace raw-modality baselines for trajectory, retrieval, reconstruction, or misalignment tasks.",
331
+ "The Cosmos3-Nano future-window branch is verified as a compatibility adapter result, Cosmos3-Super Reasoner is verified as a base-weight evaluation, and Cosmos3-Super Forward-Dynamics LoRA is verified as the first fine-tuned Super adapter branch. Cosmos3-Super adapter weights belong in cy0307/ropedia-cosmos3-super-forward-dynamics-lora-128ep; verified_public packages exclude safetensors.",
332
+ "The current reconstruction task reconstructs feature vectors, not pixel-depth, mesh, NeRF, or Gaussian reconstruction.",
333
+ "Audio is one of the synchronized source modalities in the current task representation.",
334
+ "The audio ablation report compares audio/no-audio variants across all 12 task contracts in results/audio_ablation/.",
335
+ "Foundation-model selection is explicit: Qwen3-Omni is the structured JSON baseline, Cosmos 3 is the world-model branch with Nano compatibility and Super forward-dynamics LoRA results, and policy models such as OpenVLA/openpi/GR00T wait for robot-compatible action-target conversion.",
336
+ "Future model branches should be added through the backbone registry and verified package contract, not as one-off result folders with incompatible metrics or publication rules.",
337
+ "The Xperience Embodied Foundation Model is a future native-pretraining goal, not a completed model or current benchmark."
338
+ ]
339
  }
data/public_surface_qa.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "title": "Ropedia Xperience-10M Public Project Surface",
3
  "status": "pass",
4
- "generated_at_utc": "2026-06-04T16:48:58+00:00",
5
  "scope": "Repo README, GitHub Pages HTML, Hugging Face Space card, artifact dataset card, and model card.",
6
  "checks": [
7
  {
@@ -18,7 +18,7 @@
18
  "website_integrity": {
19
  "exists": true,
20
  "status": "pass",
21
- "generated_at_utc": "2026-06-04T16:42:53+00:00"
22
  },
23
  "rendered_site_check": {
24
  "exists": true,
@@ -28,27 +28,27 @@
28
  "task_surface_integrity": {
29
  "exists": true,
30
  "status": "pass",
31
- "generated_at_utc": "2026-06-03T20:36:23+00:00"
32
  },
33
  "source_alignment": {
34
  "exists": true,
35
  "status": "pass",
36
- "generated_at_utc": "2026-06-04T16:42:48+00:00"
37
  },
38
  "scale_up_status": {
39
  "exists": true,
40
  "status": "pass",
41
- "generated_at_utc": "2026-06-04T16:42:49+00:00"
42
  },
43
  "publication_package": {
44
  "exists": true,
45
  "status": "pass",
46
- "generated_at_utc": "2026-06-04T16:42:56+00:00"
47
  },
48
  "mirror_parity": {
49
  "exists": true,
50
  "status": "pass",
51
- "generated_at_utc": "2026-06-04T16:34:37+00:00"
52
  },
53
  "live_publication": {
54
  "exists": true,
@@ -81,7 +81,7 @@
81
  "marker_counts": {
82
  "role=\"tablist\"": 3,
83
  "role=\"tab\"": 10,
84
- "role=\"tabpanel\"": 25,
85
  "aria-selected": 13,
86
  "aria-controls": 11,
87
  "moveProjectTabFocus": 2,
@@ -101,10 +101,10 @@
101
  "reason": "Public copy should consistently present the project as Ropedia Xperience-10M, with the Qwen3-Omni scale-up status.",
102
  "marker_counts": {
103
  "Ropedia Xperience-10M Task Suite": 15,
104
- "Xperience-10M": 142,
105
- "12-task": 33,
106
- "Qwen3-Omni": 108,
107
- "128-episode pilot": 11
108
  }
109
  },
110
  {
@@ -112,12 +112,12 @@
112
  "status": "pass",
113
  "reason": "Public cards should link the repo, Space, artifacts, model baselines, upstream dataset, and Ropedia dataset page.",
114
  "marker_counts": {
115
- "https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite": 66,
116
- "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite": 8,
117
- "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts": 4,
118
- "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines": 4,
119
- "https://huggingface.co/datasets/ropedia-ai/xperience-10m": 26,
120
- "https://ropedia.com/dataset": 4
121
  }
122
  },
123
  {
@@ -125,14 +125,15 @@
125
  "status": "pass",
126
  "reason": "Readers should be able to find website reference, release package, mirror, and public presentation files from public copy.",
127
  "marker_counts": {
128
- "data/project_brief.json": 9,
129
- "data/website_integrity.json": 13,
130
- "data/rendered_site_check.json": 8,
131
- "data/task_surface_integrity.json": 23,
132
- "data/publication_audit.json": 18,
133
- "data/mirror_parity.json": 12,
134
- "data/public_surface_qa.json": 12,
135
- "data/research_roadmap.json": 23
 
136
  }
137
  },
138
  {
 
1
  {
2
  "title": "Ropedia Xperience-10M Public Project Surface",
3
  "status": "pass",
4
+ "generated_at_utc": "2026-06-11T07:16:16+00:00",
5
  "scope": "Repo README, GitHub Pages HTML, Hugging Face Space card, artifact dataset card, and model card.",
6
  "checks": [
7
  {
 
18
  "website_integrity": {
19
  "exists": true,
20
  "status": "pass",
21
+ "generated_at_utc": "2026-06-11T07:11:21+00:00"
22
  },
23
  "rendered_site_check": {
24
  "exists": true,
 
28
  "task_surface_integrity": {
29
  "exists": true,
30
  "status": "pass",
31
+ "generated_at_utc": "2026-06-11T07:10:46+00:00"
32
  },
33
  "source_alignment": {
34
  "exists": true,
35
  "status": "pass",
36
+ "generated_at_utc": "2026-06-11T07:10:46+00:00"
37
  },
38
  "scale_up_status": {
39
  "exists": true,
40
  "status": "pass",
41
+ "generated_at_utc": "2026-06-11T07:10:47+00:00"
42
  },
43
  "publication_package": {
44
  "exists": true,
45
  "status": "pass",
46
+ "generated_at_utc": "2026-06-11T07:16:04+00:00"
47
  },
48
  "mirror_parity": {
49
  "exists": true,
50
  "status": "pass",
51
+ "generated_at_utc": "2026-06-11T07:10:20+00:00"
52
  },
53
  "live_publication": {
54
  "exists": true,
 
81
  "marker_counts": {
82
  "role=\"tablist\"": 3,
83
  "role=\"tab\"": 10,
84
+ "role=\"tabpanel\"": 26,
85
  "aria-selected": 13,
86
  "aria-controls": 11,
87
  "moveProjectTabFocus": 2,
 
101
  "reason": "Public copy should consistently present the project as Ropedia Xperience-10M, with the Qwen3-Omni scale-up status.",
102
  "marker_counts": {
103
  "Ropedia Xperience-10M Task Suite": 15,
104
+ "Xperience-10M": 149,
105
+ "12-task": 32,
106
+ "Qwen3-Omni": 140,
107
+ "128-episode pilot": 1
108
  }
109
  },
110
  {
 
112
  "status": "pass",
113
  "reason": "Public cards should link the repo, Space, artifacts, model baselines, upstream dataset, and Ropedia dataset page.",
114
  "marker_counts": {
115
+ "https://github.com/ChaoYue0307/ropedia-xperience-10m-task-suite": 82,
116
+ "https://huggingface.co/spaces/cy0307/ropedia-xperience-10m-task-suite": 10,
117
+ "https://huggingface.co/datasets/cy0307/ropedia-xperience-10m-task-suite-artifacts": 7,
118
+ "https://huggingface.co/cy0307/ropedia-xperience-10m-task-baselines": 10,
119
+ "https://huggingface.co/datasets/ropedia-ai/xperience-10m": 28,
120
+ "https://ropedia.com/dataset": 5
121
  }
122
  },
123
  {
 
125
  "status": "pass",
126
  "reason": "Readers should be able to find website reference, release package, mirror, and public presentation files from public copy.",
127
  "marker_counts": {
128
+ "data/project_brief.json": 8,
129
+ "data/website_integrity.json": 2,
130
+ "data/rendered_site_check.json": 2,
131
+ "data/task_surface_integrity.json": 11,
132
+ "data/publication_audit.json": 2,
133
+ "data/mirror_parity.json": 2,
134
+ "data/public_surface_qa.json": 2,
135
+ "data/research_roadmap.json": 11,
136
+ "data/task_suite_enhancement_128.json": 15
137
  }
138
  },
139
  {
data/publication_audit.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "status": "pass",
3
- "generated_at_utc": "2026-06-07T15:49:07+00:00",
4
  "checks": [
5
  {
6
  "name": "required_publication_assets_present",
@@ -52,6 +52,7 @@
52
  "PROJECT_STATUS.md": true,
53
  "RESEARCH_ROADMAP.md": true,
54
  "RESEARCH_TAKEAWAYS.md": true,
 
55
  "QUALITY_GATES.md": true,
56
  "PUBLIC_SURFACE_QA.md": true,
57
  "RENDERED_SITE_CHECK.md": true,
@@ -95,6 +96,7 @@
95
  "docs/data/task_surface_integrity.json": true,
96
  "docs/data/website_integrity.json": true,
97
  "docs/data/summary_metrics.json": true,
 
98
  "docs/assets/modalities/video.jpg": true,
99
  "docs/assets/modalities/audio.png": true,
100
  "docs/assets/modalities/depth.jpg": true,
@@ -134,7 +136,10 @@
134
  "scripts/validate_task_surface.py": true,
135
  "scripts/validate_website_integrity.py": true,
136
  "scripts/publish_hf_bundles.py": true,
137
- "scripts/omni/train_qwen3_omni_lora.py": true
 
 
 
138
  },
139
  "public_card_freshness": [
140
  {
@@ -182,8 +187,8 @@
182
  "github_repo": {
183
  "root": "repo",
184
  "exists": true,
185
- "file_count": 680,
186
- "text_file_count": 577,
187
  "largest_file": {
188
  "path": "tmp/omni_128_dataset_fetch/dataset.jsonl",
189
  "bytes": 582271586
@@ -193,8 +198,8 @@
193
  "hf_space_bundle": {
194
  "root": "hf_publish/space",
195
  "exists": true,
196
- "file_count": 582,
197
- "text_file_count": 480,
198
  "largest_file": {
199
  "path": "results/episode_task_suite/modality_reconstruction/predictions.npz",
200
  "bytes": 55702978
@@ -204,8 +209,8 @@
204
  "hf_artifact_bundle": {
205
  "root": "hf_publish/artifacts",
206
  "exists": true,
207
- "file_count": 757,
208
- "text_file_count": 631,
209
  "largest_file": {
210
  "path": "results/episode_task_suite/modality_reconstruction/predictions.npz",
211
  "bytes": 55702978
@@ -215,8 +220,8 @@
215
  "hf_model_bundle": {
216
  "root": "hf_publish/model",
217
  "exists": true,
218
- "file_count": 945,
219
- "text_file_count": 784,
220
  "largest_file": {
221
  "path": "pytorch_model.bin",
222
  "bytes": 93495480
 
1
  {
2
  "status": "pass",
3
+ "generated_at_utc": "2026-06-11T07:16:04+00:00",
4
  "checks": [
5
  {
6
  "name": "required_publication_assets_present",
 
52
  "PROJECT_STATUS.md": true,
53
  "RESEARCH_ROADMAP.md": true,
54
  "RESEARCH_TAKEAWAYS.md": true,
55
+ "TASK_SUITE_ENHANCEMENT_128.md": true,
56
  "QUALITY_GATES.md": true,
57
  "PUBLIC_SURFACE_QA.md": true,
58
  "RENDERED_SITE_CHECK.md": true,
 
96
  "docs/data/task_surface_integrity.json": true,
97
  "docs/data/website_integrity.json": true,
98
  "docs/data/summary_metrics.json": true,
99
+ "docs/data/task_suite_enhancement_128.json": true,
100
  "docs/assets/modalities/video.jpg": true,
101
  "docs/assets/modalities/audio.png": true,
102
  "docs/assets/modalities/depth.jpg": true,
 
136
  "scripts/validate_task_surface.py": true,
137
  "scripts/validate_website_integrity.py": true,
138
  "scripts/publish_hf_bundles.py": true,
139
+ "scripts/omni/build_task_suite_enhancement_128.py": true,
140
+ "scripts/omni/train_qwen3_omni_lora.py": true,
141
+ "results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json": true,
142
+ "results/omni_finetune/task_suite_enhancement_128_v1_20260608/ENHANCEMENT_REPORT.md": true
143
  },
144
  "public_card_freshness": [
145
  {
 
187
  "github_repo": {
188
  "root": "repo",
189
  "exists": true,
190
+ "file_count": 1002,
191
+ "text_file_count": 779,
192
  "largest_file": {
193
  "path": "tmp/omni_128_dataset_fetch/dataset.jsonl",
194
  "bytes": 582271586
 
198
  "hf_space_bundle": {
199
  "root": "hf_publish/space",
200
  "exists": true,
201
+ "file_count": 687,
202
+ "text_file_count": 560,
203
  "largest_file": {
204
  "path": "results/episode_task_suite/modality_reconstruction/predictions.npz",
205
  "bytes": 55702978
 
209
  "hf_artifact_bundle": {
210
  "root": "hf_publish/artifacts",
211
  "exists": true,
212
+ "file_count": 877,
213
+ "text_file_count": 726,
214
  "largest_file": {
215
  "path": "results/episode_task_suite/modality_reconstruction/predictions.npz",
216
  "bytes": 55702978
 
220
  "hf_model_bundle": {
221
  "root": "hf_publish/model",
222
  "exists": true,
223
+ "file_count": 1067,
224
+ "text_file_count": 881,
225
  "largest_file": {
226
  "path": "pytorch_model.bin",
227
  "bytes": 93495480
data/quality_gates.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "title": "Ropedia Xperience-10M Release Checks",
3
  "status": "pass",
4
- "generated_at_utc": "2026-06-04T07:39:12+00:00",
5
  "rule": "A release is current when the automated reports pass and the live GitHub/Hugging Face mirrors are verified after publishing.",
6
  "automated_gates": [
7
  {
@@ -154,7 +154,7 @@
154
  "command": "python scripts/validate_mirror_parity.py",
155
  "report": "docs/data/mirror_parity.json",
156
  "blocks_if": "Prepared HF Space, artifact dataset, or model bundle diverges from the repo for critical files.",
157
- "shows": "The files prepared for GitHub and Hugging Face are synchronized before upload.",
158
  "current_report": {
159
  "exists": true,
160
  "status": "pass"
 
1
  {
2
  "title": "Ropedia Xperience-10M Release Checks",
3
  "status": "pass",
4
+ "generated_at_utc": "2026-06-11T07:16:22+00:00",
5
  "rule": "A release is current when the automated reports pass and the live GitHub/Hugging Face mirrors are verified after publishing.",
6
  "automated_gates": [
7
  {
 
154
  "command": "python scripts/validate_mirror_parity.py",
155
  "report": "docs/data/mirror_parity.json",
156
  "blocks_if": "Prepared HF Space, artifact dataset, or model bundle diverges from the repo for critical files.",
157
+ "shows": "The files staged for GitHub and Hugging Face are synchronized before upload.",
158
  "current_report": {
159
  "exists": true,
160
  "status": "pass"
data/qwen3_full_parameter_gates.json ADDED
@@ -0,0 +1,219 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "title": "Qwen3-Omni Full-Parameter Feasibility Gates",
3
+ "generated_at_utc": "2026-06-11T03:21:58+00:00",
4
+ "status": "pass",
5
+ "decision": "full_parameter_feasible_for_guarded_short_runs_not_promoted",
6
+ "interpretation": "The 2026-06-09 gates prove that Qwen3-Omni full-parameter FSDP can load, prepare, run backward/optimizer steps, and complete guarded pilots up to 128 optimizer steps on an 8-GPU remote worker. They do not prove a production full-parameter fine-tune, and they intentionally save no full checkpoints or public weights.",
7
+ "aggregate": {
8
+ "run_count": 6,
9
+ "passed_run_count": 5,
10
+ "preempted_run_count": 1,
11
+ "review_or_missing_run_count": 0,
12
+ "completed_full_parameter_train_steps": 233,
13
+ "longest_passed_run_id": "xperience10m_qwen3_omni_128ep_fullparam_pilot128_after_qwen_v5_preemptible_8gpu_20260609",
14
+ "longest_passed_steps": 128,
15
+ "num_processes": [
16
+ 8
17
+ ],
18
+ "checkpoint_saved": false
19
+ },
20
+ "runs": [
21
+ {
22
+ "id": "fullparam_smoke_1step",
23
+ "title": "Full-Parameter 1-Step Feasibility Smoke",
24
+ "status": "passed",
25
+ "scope": "1 optimizer step over 8 train samples",
26
+ "summary_path": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_smoke_preemptible_8gpu_20260609/fullparam_feasibility_summary.json",
27
+ "run_id": "xperience10m_qwen3_omni_128ep_fullparam_smoke_preemptible_8gpu_20260609",
28
+ "purpose": "Full-parameter Qwen3-Omni feasibility gate: load, FSDP prepare, backward/optimizer step, and no checkpoint save.",
29
+ "tuning_mode": "full",
30
+ "training_objective": null,
31
+ "num_processes": 8,
32
+ "num_train_samples": 8,
33
+ "configured_max_train_steps": 1,
34
+ "observed_train_steps": 1,
35
+ "first_step_loss": null,
36
+ "final_step_loss": 1.2726006507873535,
37
+ "epoch_train_loss": 1.2726006507873535,
38
+ "min_step_loss": null,
39
+ "max_step_loss": null,
40
+ "model_load_seconds": null,
41
+ "accelerator_prepare_seconds": null,
42
+ "train_loop_seconds": null,
43
+ "save_mode": "none",
44
+ "checkpoint_saved": false,
45
+ "checkpoint_policy": "no full-parameter checkpoint or public weights; save_mode=none",
46
+ "preempt_event": null,
47
+ "parent_resume_event": null
48
+ },
49
+ {
50
+ "id": "fullparam_shorttrain8",
51
+ "title": "Full-Parameter 8-Step Short Train",
52
+ "status": "passed",
53
+ "scope": "8 optimizer steps over 64 train samples",
54
+ "summary_path": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_shorttrain8_preemptible_8gpu_20260609/fullparam_shorttrain8_summary.json",
55
+ "run_id": "xperience10m_qwen3_omni_128ep_fullparam_shorttrain8_preemptible_8gpu_20260609",
56
+ "purpose": "guarded full-parameter Qwen3-Omni short train using all 8 remote GPUs while Qwen v5 export was CPU-bound",
57
+ "tuning_mode": "full",
58
+ "training_objective": "structured_episode_understanding_json_qa",
59
+ "num_processes": 8,
60
+ "num_train_samples": 64,
61
+ "configured_max_train_steps": 8,
62
+ "observed_train_steps": 8,
63
+ "first_step_loss": 1.2701153755187988,
64
+ "final_step_loss": 1.180522084236145,
65
+ "epoch_train_loss": 1.2190196067094803,
66
+ "min_step_loss": 1.1532728672027588,
67
+ "max_step_loss": 1.2701153755187988,
68
+ "model_load_seconds": 111.21318578720093,
69
+ "accelerator_prepare_seconds": null,
70
+ "train_loop_seconds": null,
71
+ "save_mode": "none",
72
+ "checkpoint_saved": false,
73
+ "checkpoint_policy": "no full-parameter checkpoint or public weights; save_mode=none",
74
+ "preempt_event": null,
75
+ "parent_resume_event": null
76
+ },
77
+ {
78
+ "id": "fullparam_pilot32",
79
+ "title": "Full-Parameter 32-Step Pilot",
80
+ "status": "passed",
81
+ "scope": "32 optimizer steps over 256 train samples",
82
+ "summary_path": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot32_preemptible_8gpu_20260609/fullparam_pilot32_summary.json",
83
+ "run_id": "xperience10m_qwen3_omni_128ep_fullparam_pilot32_preemptible_8gpu_20260609",
84
+ "purpose": "guarded full-parameter Qwen3-Omni 32-step pilot using all 8 remote GPUs while Qwen v5 export was CPU-bound",
85
+ "tuning_mode": "full",
86
+ "training_objective": "structured_episode_understanding_json_qa",
87
+ "num_processes": 8,
88
+ "num_train_samples": 256,
89
+ "configured_max_train_steps": 32,
90
+ "observed_train_steps": 32,
91
+ "first_step_loss": 1.2583622932434082,
92
+ "final_step_loss": 0.2206273376941681,
93
+ "epoch_train_loss": 0.8451133379712701,
94
+ "min_step_loss": 0.2206273376941681,
95
+ "max_step_loss": 1.26246976852417,
96
+ "model_load_seconds": 116.43117642402649,
97
+ "accelerator_prepare_seconds": 1122.333820104599,
98
+ "train_loop_seconds": 440.2827479839325,
99
+ "save_mode": "none",
100
+ "checkpoint_saved": false,
101
+ "checkpoint_policy": "no full-parameter checkpoint or public weights; save_mode=none",
102
+ "preempt_event": null,
103
+ "parent_resume_event": null
104
+ },
105
+ {
106
+ "id": "fullparam_pilot64",
107
+ "title": "Full-Parameter 64-Step Pilot",
108
+ "status": "passed",
109
+ "scope": "64 optimizer steps over 512 train samples",
110
+ "summary_path": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot64_preemptible_8gpu_20260609/fullparam_pilot64_summary.json",
111
+ "run_id": "xperience10m_qwen3_omni_128ep_fullparam_pilot64_preemptible_8gpu_20260609",
112
+ "purpose": "guarded full-parameter Qwen3-Omni 64-step pilot using all 8 remote GPUs while Qwen v5 long multiscale export was CPU-bound",
113
+ "tuning_mode": "full",
114
+ "training_objective": "structured_episode_understanding_json_qa",
115
+ "num_processes": 8,
116
+ "num_train_samples": 512,
117
+ "configured_max_train_steps": 64,
118
+ "observed_train_steps": 64,
119
+ "first_step_loss": 1.2597532272338867,
120
+ "final_step_loss": 0.011219973675906658,
121
+ "epoch_train_loss": 0.4434075650788145,
122
+ "min_step_loss": 0.011219973675906658,
123
+ "max_step_loss": 1.2597532272338867,
124
+ "model_load_seconds": 80.75278043746948,
125
+ "accelerator_prepare_seconds": 811.6013672351837,
126
+ "train_loop_seconds": 903.8602497577667,
127
+ "save_mode": "none",
128
+ "checkpoint_saved": false,
129
+ "checkpoint_policy": "no full-parameter checkpoint or public weights; save_mode=none",
130
+ "preempt_event": null,
131
+ "parent_resume_event": null
132
+ },
133
+ {
134
+ "id": "fullparam_pilot128_preempted",
135
+ "title": "Full-Parameter 128-Step Opportunistic Pilot",
136
+ "status": "preempted_for_qwen_v5_handoff",
137
+ "scope": "planned 128 optimizer steps over 1024 train samples; preempted for Qwen v5 handoff",
138
+ "summary_path": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot128_preemptible_8gpu_20260609/fullparam_pilot128_summary.json",
139
+ "run_id": "xperience10m_qwen3_omni_128ep_fullparam_pilot128_preemptible_8gpu_20260609",
140
+ "purpose": "opportunistic guarded full-parameter Qwen3-Omni 128-step pilot launched only while Qwen v5 export was CPU-bound; preempted for main Qwen v5 handoff",
141
+ "tuning_mode": "full",
142
+ "training_objective": "structured_episode_understanding_json_qa",
143
+ "num_processes": 8,
144
+ "num_train_samples": 1024,
145
+ "configured_max_train_steps": 128,
146
+ "observed_train_steps": 0,
147
+ "first_step_loss": null,
148
+ "final_step_loss": null,
149
+ "epoch_train_loss": null,
150
+ "min_step_loss": null,
151
+ "max_step_loss": null,
152
+ "model_load_seconds": 65.33229899406433,
153
+ "accelerator_prepare_seconds": null,
154
+ "train_loop_seconds": null,
155
+ "save_mode": "none",
156
+ "checkpoint_saved": false,
157
+ "checkpoint_policy": "no full-parameter checkpoint or public weights; save_mode=none",
158
+ "preempt_event": {
159
+ "event": "fullparam_pilot128_preempt_for_qwen_v5_handoff",
160
+ "guard_pid": 3757690,
161
+ "kind": "fullparam_pilot128_handoff_guard",
162
+ "time": 1780954910,
163
+ "watched_pid": 3753782
164
+ },
165
+ "parent_resume_event": {
166
+ "event": "fullparam_pilot128_resumed_parent_after_preempt",
167
+ "guard_pid": 3757690,
168
+ "kind": "fullparam_pilot128_handoff_guard",
169
+ "time": 1780954925,
170
+ "watched_pid": 3753782
171
+ }
172
+ },
173
+ {
174
+ "id": "fullparam_pilot128_after_qwen_v5",
175
+ "title": "Full-Parameter 128-Step Post-Qwen-v5 Pilot",
176
+ "status": "passed",
177
+ "scope": "128 optimizer steps over 1024 train samples after verified Qwen v5 handoff",
178
+ "summary_path": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot128_after_qwen_v5_preemptible_8gpu_20260609/training_metadata.json",
179
+ "progress_path": "results/omni_finetune/xperience10m_qwen3_omni_128ep_fullparam_pilot128_after_qwen_v5_preemptible_8gpu_20260609/progress.jsonl",
180
+ "run_id": "xperience10m_qwen3_omni_128ep_fullparam_pilot128_after_qwen_v5_preemptible_8gpu_20260609",
181
+ "purpose": "post_verified_qwen_v5_full_parameter_feasibility_pilot",
182
+ "tuning_mode": "full",
183
+ "training_objective": "structured_episode_understanding_json_qa",
184
+ "num_processes": 8,
185
+ "num_train_samples": 1024,
186
+ "configured_max_train_steps": 128,
187
+ "observed_train_steps": 128,
188
+ "first_step_loss": 1.2273844480514526,
189
+ "final_step_loss": 0.0136940386146307,
190
+ "epoch_train_loss": 0.21579630990163423,
191
+ "min_step_loss": 0.004702376667410135,
192
+ "max_step_loss": 1.2273844480514526,
193
+ "model_load_seconds": null,
194
+ "accelerator_prepare_seconds": null,
195
+ "train_loop_seconds": null,
196
+ "save_mode": "none",
197
+ "checkpoint_saved": false,
198
+ "checkpoint_policy": "no full-parameter checkpoint or public weights; save_mode=none",
199
+ "preempt_event": null,
200
+ "parent_resume_event": null,
201
+ "progress_events": {
202
+ "max_steps_reached": true,
203
+ "save_skipped": true,
204
+ "complete": true
205
+ }
206
+ }
207
+ ],
208
+ "publication_policy": {
209
+ "public_summary_allowed": true,
210
+ "publish_full_parameter_weights": false,
211
+ "publish_full_checkpoints": false,
212
+ "reason": "All completed 2026-06-09 full-parameter runs used save_mode=none; the preempted pilot saved nothing. These are feasibility evidence only."
213
+ },
214
+ "next_steps": [
215
+ "Keep the verified Qwen3-Omni LoRA adapter as the published production result for the 128-episode suite.",
216
+ "For a production full-parameter run, add a sharded checkpoint/resume plan before any long training launch.",
217
+ "Run a separate checkpointed full-parameter pilot only when GPUs are not needed by verified LoRA evaluation/publication work."
218
+ ]
219
+ }
data/research_roadmap.json CHANGED
@@ -1,217 +1,241 @@
1
  {
2
- "title": "Ropedia Xperience-10M Research Roadmap",
3
- "summary": "Staged path from the public-sample task lab to a final verified Qwen3-Omni diagnostic result, same-split 128-episode baseline alignment, action/subtask error analysis, foundation-model selection, world/policy branches, and a future Xperience-native embodied foundation model.",
4
- "current_decision_point": "Keep the public-sample task suite as the development harness, use the final verified selected-episode Qwen3-Omni diagnostic result and the same-split 128-episode simple/NN metadata baselines as the first cross-episode references, improve action/subtask quality through error analysis, then branch into Cosmos 3 world modeling and policy-model experiments after their targets are implemented. The Xperience Embodied Foundation Model is a later full-corpus pretraining goal, not a current result.",
5
- "additional_development_directions": {
6
- "source_document": "ADDITIONAL_DEVELOPMENT_DIRECTIONS.md",
7
- "source_json": "docs/data/additional_development_directions.json",
8
- "summary": "Additional concrete tracks include episode taxonomy and data selection, benchmark protocol, multimodal representation learning, skill graphs, affordance modeling, 3D/4D scene memory, data-quality diagnostics, and policy/simulation transfer."
9
- },
10
- "phases": [
11
- {
12
- "id": "public_sample_task_lab",
13
- "name": "Public-Sample Task Lab",
14
- "status": "implemented",
15
- "entry_condition": "One public Xperience-10M sample episode is available.",
16
- "deliverables": [
17
- "1161 aligned windows",
18
- "12 task contracts",
19
- "minimal baseline heads",
20
- "neural MLP heads",
21
- "modality atlas",
22
- "task walkthroughs",
23
- "derived figures"
24
- ],
25
- "completion_evidence": [
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
26
  "PROJECT_STATUS.md",
27
- "EVALUATION_PROTOCOL.md",
28
  "RESEARCH_TAKEAWAYS.md",
29
- "docs/data/summary_metrics.json",
30
- "results/episode_task_suite/summary_report.json"
31
- ],
32
- "reader_takeaway": "The public sample supports task design, feature contracts, walkthroughs, and baseline comparisons."
33
- },
34
- {
35
- "id": "multi_episode_data_staging",
36
- "name": "Multi-Episode Data Preparation",
37
- "status": "implemented_for_first_pilot",
38
- "entry_condition": "Gated dataset availability and enough storage for selected episodes.",
39
- "deliverables": [
40
- "128 selected episodes",
41
- "episode manifest",
42
- "missing-view manifest",
43
- "held-out episode split",
44
- "source-discovery report"
45
- ],
46
- "completion_evidence": [
47
- "results/omni_finetune/DATA_ACCESS_STATUS.md",
48
- "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md",
49
- "results/omni_finetune/source_discovery.json"
50
- ],
51
- "reader_takeaway": "The first selected split is available for Qwen3-Omni diagnostics, with train/test separation at the episode level."
52
- },
53
- {
54
- "id": "qwen3_omni_lora_diagnostic_pilot",
55
- "name": "Qwen3-Omni LoRA Final Diagnostic Result",
56
- "status": "verified_baseline",
57
- "entry_condition": "Selected episodes are prepared locally with no train/test episode leakage.",
58
- "deliverables": [
59
- "dataset JSONL/media manifests",
60
- "LoRA adapter checkpoint",
61
- "progress logs",
62
- "validation monitoring",
63
- "held-out predictions",
64
- "metrics",
65
- "confusion matrices",
66
- "run report",
67
- "public LoRA adapter repo"
68
- ],
69
- "completion_evidence": [
70
- "docs/data/omni_finetune_verified_result.json",
71
- "results/omni_finetune/verified_public/",
72
- "dataset_manifest.json",
73
- "training_metadata.json",
74
- "progress.jsonl",
75
- "metrics.json",
76
- "predictions.jsonl",
77
- "RUN_REPORT.md"
78
- ],
79
- "reader_takeaway": "The final omni-model diagnostic result establishes the full held-out training/validation/evaluation loop and meets the strict-JSON target, but weak action/subtask metrics make it a diagnostic baseline."
80
- },
81
- {
82
- "id": "multi_episode_128_same_split_baselines",
83
- "name": "128-Episode Same-Split Simple/NN Baselines",
84
- "status": "verified_companion_result",
85
- "entry_condition": "Derived Qwen JSONL export for the selected 96/16/16 split.",
86
- "deliverables": [
87
- "same 12 task ids",
88
- "simple metadata/text baselines",
89
- "neural MLP baselines for JSON-supported labels",
90
- "explicit unsupported markers for raw-feature-only tasks"
91
- ],
92
- "completion_evidence": [
93
- "results/omni_finetune/multi_episode_128_task_baselines/BASELINE_ALIGNMENT_REPORT.md",
94
- "results/omni_finetune/multi_episode_128_task_baselines/summary_report.json",
95
- "scripts/omni/run_128_task_baselines.py"
96
- ],
97
- "reader_takeaway": "The simple and neural baseline framing is now aligned to the selected 128-episode setup; trajectory, retrieval, reconstruction, and misalignment variants still need raw 128 feature blocks for exact feature-level reproduction."
98
- },
99
- {
100
- "id": "qwen3_omni_structured_output_error_analysis",
101
- "name": "Action/Subtask Error-Analysis Pass",
102
- "status": "active_next_step",
103
- "entry_condition": "The final diagnostic package meets strict JSON validity but has weak action/subtask held-out quality.",
104
- "deliverables": [
105
- "same 96/16/16 episode split",
106
- "action/subtask confusion analysis",
107
- "unseen-label analysis",
108
- "object/action family breakdowns",
109
- "held-out test evaluation",
110
- "comparison to the final verified Qwen baseline"
111
- ],
112
- "completion_evidence": [
113
- "error-analysis tables",
114
- "held-out metrics by failure type",
115
- "verified public-safe package"
116
- ],
117
- "reader_takeaway": "The next pass should improve action/subtask quality before larger model-quality claims."
118
- },
119
- {
120
- "id": "foundation_model_selection_matrix",
121
- "name": "Foundation-Model Selection Matrix",
122
- "status": "next",
123
- "entry_condition": "The selected episodes are prepared or a 3-8 episode dry run is available for preprocessing checks.",
124
- "deliverables": [
125
- "backbone registry",
126
- "Cosmos 3 world-model branch plan",
127
- "Qwen3-Omni LoRA baseline plan",
128
- "OpenVLA/openpi/GR00T policy-branch candidates",
129
- "model-specific evaluation additions"
130
- ],
131
- "completion_evidence": [
132
- "FOUNDATION_MODEL_PLAN.md",
133
- "docs/data/foundation_model_plan.json",
134
- "research_roadmap_interactive.json"
135
- ],
136
- "reader_takeaway": "Qwen3-Omni remains the first trainable held-out pilot; Cosmos 3 is the first world-model branch. Cosmos3-Super now has camera-pose proxy forward-dynamics targets ready for trainer implementation, while VLA/policy models wait for robot-compatible action targets."
137
- },
138
- {
139
- "id": "robustness_run_64_128_episode",
140
- "name": "64-128 Episode Robustness Run",
141
- "status": "planned",
142
- "entry_condition": "The selected-episode pilot trains and evaluates cleanly.",
143
- "deliverables": [
144
- "split-by-session metrics",
145
- "modality ablations",
146
- "calibration/object/language error analysis",
147
- "missing-view sensitivity analysis"
148
- ],
149
- "completion_evidence": [
150
- "held-out metrics by session",
151
- "held-out metrics by task",
152
- "held-out metrics by modality",
153
- "ablation tables",
154
- "qualitative error analysis"
155
- ],
156
- "reader_takeaway": "The robustness run tests whether the pilot conclusions survive broader sessions and missing modalities."
157
- },
158
- {
159
- "id": "foundation_world_model_extensions",
160
- "name": "Cosmos 3 and Policy-Model Extensions",
161
- "status": "planned",
162
- "entry_condition": "Enough multi-episode data, compute budget, and model-specific action/world-state targets.",
163
- "deliverables": [
164
- "Cosmos 3 future-window or action-conditioned world-model probe",
165
- "OpenVLA/openpi/GR00T action-policy baseline",
166
- "audio/video/depth/pose/mocap conditioning checks",
167
- "affordance and object-interaction tasks",
168
- "synthetic-data usefulness test"
169
- ],
170
- "completion_evidence": [
171
- "task-specific held-out evaluations",
172
- "qualitative inspection",
173
- "updated model cards"
174
- ],
175
- "reader_takeaway": "The long-term direction is richer multimodal representation learning for embodied-AI reasoning, with model branches chosen by task fit rather than by a single default backbone."
176
- },
177
- {
178
- "id": "xperience_embodied_foundation_pretraining",
179
- "name": "Xperience Embodied Foundation Model Pretraining",
180
- "status": "future",
181
- "entry_condition": "Full-corpus access, PB-scale storage path, high-throughput data loading, multi-node compute, and positive scaling evidence from smaller multi-episode runs.",
182
- "deliverables": [
183
- "full-corpus episode and split manifests",
184
- "pretraining shard and provenance manifests",
185
- "0.3B-1B and 1B-3B scaling pilots",
186
- "3B-7B Xperience-native domain model target",
187
- "held-out episode/session/activity/object evaluations",
188
- "missing-modality robustness report",
189
- "model card and data-boundary report"
190
- ],
191
- "completion_evidence": [
192
- "pretraining metadata",
193
- "checkpoint inventory",
194
- "scaling curves",
195
- "held-out evaluation reports",
196
- "qualitative retrieval or future-state examples",
197
- "safety and data-boundary report"
198
- ],
199
- "reader_takeaway": "The final research direction is a domain-specific embodied foundation model trained directly on Xperience-10M, after smaller pilots justify the cost and infrastructure."
200
- }
201
- ],
202
- "public_surfaces_to_update": [
203
- "README.md",
204
- "PROJECT_STATUS.md",
205
- "RESEARCH_TAKEAWAYS.md",
206
- "EVALUATION_PROTOCOL.md",
207
- "ARTIFACT_GUIDE.md",
208
- "ADDITIONAL_DEVELOPMENT_DIRECTIONS.md",
209
- "XPERIENCE_EMBODIED_FOUNDATION_MODEL_PRETRAINING.md",
210
- "docs/index.html",
211
- "docs/data/additional_development_directions.json",
212
- "docs/data/research_roadmap.json",
213
- "Hugging Face Space card",
214
- "Hugging Face artifact dataset card",
215
- "Hugging Face model card"
216
- ]
217
  }
 
1
  {
2
+ "title": "Ropedia Xperience-10M Research Roadmap",
3
+ "summary": "Staged path from the public-sample task lab to verified Qwen3-Omni, Cosmos3-Nano, and Cosmos3-Super diagnostics, same-split 128-episode baseline alignment, a no-new-episode 128-suite enhancement pack, action/subtask error analysis, world/policy branches, and a future Xperience-native embodied foundation model.",
4
+ "current_decision_point": "Push the current selected 128 episodes harder before requesting more storage: keep the public-sample task suite as the development harness, use the final verified selected-episode Qwen3-Omni v4 diagnostic result and same-split 128-episode simple/NN metadata baselines as structured-task references, read Cosmos3-Nano and Cosmos3-Super Forward-Dynamics LoRA as separate world-model results, export multiscale_20s10_40s20_80s40 windows plus hierarchical action/subtask targets, and defer policy-model experiments until robot-compatible targets are implemented. The Xperience Embodied Foundation Model is a later full-corpus pretraining goal, not a current result.",
5
+ "additional_development_directions": {
6
+ "source_document": "ADDITIONAL_DEVELOPMENT_DIRECTIONS.md",
7
+ "source_json": "docs/data/additional_development_directions.json",
8
+ "summary": "Additional concrete tracks include episode taxonomy and data selection, benchmark protocol, multimodal representation learning, skill graphs, affordance modeling, 3D/4D scene memory, data-quality diagnostics, and policy/simulation transfer."
9
+ },
10
+ "phases": [
11
+ {
12
+ "id": "public_sample_task_lab",
13
+ "name": "Public-Sample Task Lab",
14
+ "status": "implemented",
15
+ "entry_condition": "One public Xperience-10M sample episode is available.",
16
+ "deliverables": [
17
+ "1161 aligned windows",
18
+ "12 task contracts",
19
+ "minimal baseline heads",
20
+ "neural MLP heads",
21
+ "modality atlas",
22
+ "task walkthroughs",
23
+ "derived figures"
24
+ ],
25
+ "completion_evidence": [
26
+ "PROJECT_STATUS.md",
27
+ "EVALUATION_PROTOCOL.md",
28
+ "RESEARCH_TAKEAWAYS.md",
29
+ "docs/data/summary_metrics.json",
30
+ "results/episode_task_suite/summary_report.json"
31
+ ],
32
+ "reader_takeaway": "The public sample supports task design, feature contracts, walkthroughs, and baseline comparisons."
33
+ },
34
+ {
35
+ "id": "multi_episode_data_staging",
36
+ "name": "Multi-Episode Data Preparation",
37
+ "status": "implemented_for_first_pilot",
38
+ "entry_condition": "Gated dataset availability and enough storage for selected episodes.",
39
+ "deliverables": [
40
+ "128 selected episodes",
41
+ "episode manifest",
42
+ "missing-view manifest",
43
+ "held-out episode split",
44
+ "source-discovery report"
45
+ ],
46
+ "completion_evidence": [
47
+ "results/omni_finetune/DATA_ACCESS_STATUS.md",
48
+ "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md",
49
+ "results/omni_finetune/source_discovery.json"
50
+ ],
51
+ "reader_takeaway": "The first selected split is available for Qwen3-Omni diagnostics, with train/test separation at the episode level."
52
+ },
53
+ {
54
+ "id": "qwen3_omni_lora_diagnostic_pilot",
55
+ "name": "Qwen3-Omni LoRA Final Diagnostic Result",
56
+ "status": "verified_baseline",
57
+ "entry_condition": "Selected episodes are prepared locally with no train/test episode leakage.",
58
+ "deliverables": [
59
+ "dataset JSONL/media manifests",
60
+ "LoRA adapter checkpoint",
61
+ "progress logs",
62
+ "validation monitoring",
63
+ "held-out predictions",
64
+ "metrics",
65
+ "confusion matrices",
66
+ "run report",
67
+ "public LoRA adapter repo"
68
+ ],
69
+ "completion_evidence": [
70
+ "docs/data/omni_finetune_verified_result.json",
71
+ "results/omni_finetune/verified_public/",
72
+ "dataset_manifest.json",
73
+ "training_metadata.json",
74
+ "progress.jsonl",
75
+ "metrics.json",
76
+ "predictions.jsonl",
77
+ "RUN_REPORT.md"
78
+ ],
79
+ "reader_takeaway": "The final omni-model diagnostic result establishes the full held-out training/validation/evaluation loop and meets the strict-JSON target, but weak action/subtask metrics make it a diagnostic baseline."
80
+ },
81
+ {
82
+ "id": "multi_episode_128_same_split_baselines",
83
+ "name": "128-Episode Same-Split Simple/NN Baselines",
84
+ "status": "verified_companion_result",
85
+ "entry_condition": "Derived Qwen JSONL export for the selected 96/16/16 split.",
86
+ "deliverables": [
87
+ "same 12 task ids",
88
+ "simple metadata/text baselines",
89
+ "neural MLP baselines for JSON-supported labels",
90
+ "explicit unsupported markers for raw-feature-only tasks"
91
+ ],
92
+ "completion_evidence": [
93
+ "results/omni_finetune/multi_episode_128_task_baselines/BASELINE_ALIGNMENT_REPORT.md",
94
+ "results/omni_finetune/multi_episode_128_task_baselines/summary_report.json",
95
+ "scripts/omni/run_128_task_baselines.py"
96
+ ],
97
+ "reader_takeaway": "The simple and neural baseline framing is now aligned to the selected 128-episode setup; trajectory, retrieval, reconstruction, and misalignment variants still need raw 128 feature blocks for exact feature-level reproduction."
98
+ },
99
+ {
100
+ "id": "task_suite_enhancement_128",
101
+ "name": "128-Episode Task Suite Enhancement Pack",
102
+ "status": "current",
103
+ "entry_condition": "Same selected 96/16/16 split and current public 3,808-window export.",
104
+ "deliverables": [
105
+ "dense-window and multiscale export estimates",
106
+ "hierarchical action/subtask target contract",
107
+ "raw-feature shard priorities for unsupported tasks",
108
+ "Qwen v5 and Cosmos continuation run cards",
109
+ "publication-ready enhancement artifacts"
110
+ ],
111
+ "completion_evidence": [
112
+ "TASK_SUITE_ENHANCEMENT_128.md",
113
+ "docs/data/task_suite_enhancement_128.json",
114
+ "results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json",
115
+ "scripts/omni/build_task_suite_enhancement_128.py"
116
+ ],
117
+ "reader_takeaway": "The current 128-episode setup still has headroom: use multiscale_20s10_40s20_80s40, hierarchical labels, label-normalized scoring, and raw-feature shards before adding more episodes."
118
+ },
119
+ {
120
+ "id": "qwen3_omni_structured_output_error_analysis",
121
+ "name": "Action/Subtask Error-Analysis Pass",
122
+ "status": "active_next_step",
123
+ "entry_condition": "The final diagnostic package meets strict JSON validity but has weak action/subtask held-out quality.",
124
+ "deliverables": [
125
+ "same 96/16/16 episode split",
126
+ "action/subtask confusion analysis",
127
+ "unseen-label analysis",
128
+ "object/action family breakdowns",
129
+ "held-out test evaluation",
130
+ "comparison to the final verified Qwen baseline"
131
+ ],
132
+ "completion_evidence": [
133
+ "error-analysis tables",
134
+ "held-out metrics by failure type",
135
+ "verified public-safe package"
136
+ ],
137
+ "reader_takeaway": "The next pass should improve action/subtask quality before larger model-quality claims."
138
+ },
139
+ {
140
+ "id": "foundation_model_selection_matrix",
141
+ "name": "Foundation-Model Selection Matrix",
142
+ "status": "current",
143
+ "entry_condition": "The selected episodes are prepared or a 3-8 episode dry run is available for preprocessing checks.",
144
+ "deliverables": [
145
+ "backbone registry",
146
+ "Cosmos 3 world-model branch plan",
147
+ "Cosmos3-Super Forward-Dynamics LoRA verified package",
148
+ "Qwen3-Omni LoRA baseline plan",
149
+ "OpenVLA/openpi/GR00T policy-branch candidates",
150
+ "model-specific evaluation additions"
151
+ ],
152
+ "completion_evidence": [
153
+ "FOUNDATION_MODEL_PLAN.md",
154
+ "docs/data/foundation_model_plan.json",
155
+ "research_roadmap_interactive.json"
156
+ ],
157
+ "reader_takeaway": "Qwen3-Omni remains the structured JSON held-out pilot; Cosmos 3 is the first world-model branch. Cosmos3-Super now has a verified forward-dynamics LoRA over camera-pose proxy targets, while VLA/policy models wait for robot-compatible action targets."
158
+ },
159
+ {
160
+ "id": "robustness_run_64_128_episode",
161
+ "name": "64-128 Episode Robustness Run",
162
+ "status": "partially_implemented",
163
+ "entry_condition": "The selected-episode pilot trains and evaluates cleanly.",
164
+ "deliverables": [
165
+ "split-by-session metrics",
166
+ "modality ablations",
167
+ "calibration/object/language error analysis",
168
+ "missing-view sensitivity analysis"
169
+ ],
170
+ "completion_evidence": [
171
+ "held-out metrics by session",
172
+ "held-out metrics by task",
173
+ "held-out metrics by modality",
174
+ "ablation tables",
175
+ "qualitative error analysis"
176
+ ],
177
+ "reader_takeaway": "The robustness run tests whether the pilot conclusions survive broader sessions and missing modalities."
178
+ },
179
+ {
180
+ "id": "foundation_world_model_extensions",
181
+ "name": "Cosmos 3 and Policy-Model Extensions",
182
+ "status": "planned",
183
+ "entry_condition": "Enough multi-episode data, compute budget, and model-specific action/world-state targets.",
184
+ "deliverables": [
185
+ "Cosmos 3 future-window and action-conditioned world-model probes",
186
+ "OpenVLA/openpi/GR00T action-policy baseline",
187
+ "audio/video/depth/pose/mocap conditioning checks",
188
+ "affordance and object-interaction tasks",
189
+ "synthetic-data usefulness test"
190
+ ],
191
+ "completion_evidence": [
192
+ "task-specific held-out evaluations",
193
+ "verified Cosmos3-Super forward-dynamics LoRA package",
194
+ "qualitative inspection",
195
+ "updated model cards"
196
+ ],
197
+ "reader_takeaway": "The Cosmos branch now includes Nano future-window compatibility and Super forward-dynamics LoRA; the long-term direction remains richer multimodal representation learning with model branches chosen by task fit rather than by a single default backbone."
198
+ },
199
+ {
200
+ "id": "xperience_embodied_foundation_pretraining",
201
+ "name": "Xperience Embodied Foundation Model Pretraining",
202
+ "status": "future",
203
+ "entry_condition": "Full-corpus access, PB-scale storage path, high-throughput data loading, multi-node compute, and positive scaling evidence from smaller multi-episode runs.",
204
+ "deliverables": [
205
+ "full-corpus episode and split manifests",
206
+ "pretraining shard and provenance manifests",
207
+ "0.3B-1B and 1B-3B scaling pilots",
208
+ "3B-7B Xperience-native domain model target",
209
+ "held-out episode/session/activity/object evaluations",
210
+ "missing-modality robustness report",
211
+ "model card and data-boundary report"
212
+ ],
213
+ "completion_evidence": [
214
+ "pretraining metadata",
215
+ "checkpoint inventory",
216
+ "scaling curves",
217
+ "held-out evaluation reports",
218
+ "qualitative retrieval or future-state examples",
219
+ "safety and data-boundary report"
220
+ ],
221
+ "reader_takeaway": "The final research direction is a domain-specific embodied foundation model trained directly on Xperience-10M, after smaller pilots justify the cost and infrastructure."
222
+ }
223
+ ],
224
+ "public_surfaces_to_update": [
225
+ "README.md",
226
+ "docs/data/task_suite_enhancement_128.json",
227
+ "TASK_SUITE_ENHANCEMENT_128.md",
228
  "PROJECT_STATUS.md",
 
229
  "RESEARCH_TAKEAWAYS.md",
230
+ "EVALUATION_PROTOCOL.md",
231
+ "ARTIFACT_GUIDE.md",
232
+ "ADDITIONAL_DEVELOPMENT_DIRECTIONS.md",
233
+ "XPERIENCE_EMBODIED_FOUNDATION_MODEL_PRETRAINING.md",
234
+ "docs/index.html",
235
+ "docs/data/additional_development_directions.json",
236
+ "docs/data/research_roadmap.json",
237
+ "Hugging Face Space card",
238
+ "Hugging Face artifact dataset card",
239
+ "Hugging Face model card"
240
+ ]
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
241
  }
data/research_roadmap_interactive.json CHANGED
@@ -2040,12 +2040,12 @@
2040
  "step": 2
2041
  },
2042
  {
2043
- "action": "Run 3-8 episode dry runs for Qwen3-Omni prompt/LoRA, Cosmos 3 preprocessing, and one policy candidate.",
2044
  "name": "Model-selection dry run",
2045
  "step": 3
2046
  },
2047
  {
2048
- "action": "Promote Cosmos 3 if future-window/action-conditioned preprocessing fits storage and compute.",
2049
  "name": "World-model branch",
2050
  "step": 4
2051
  },
@@ -2084,8 +2084,8 @@
2084
  {
2085
  "best_role": "Embodied world modeling, action generation, future-window prediction, and synthetic-data expansion.",
2086
  "category": "world_foundation_model",
2087
- "current_decision": "add_as_first_world_model_branch_after_data_gate",
2088
- "entry_condition": "Multi-episode data plus enough storage/compute for generated or latent video-state outputs.",
2089
  "family": "Cosmos 3",
2090
  "openness": "track_official_nvidia_release_and_available_weights",
2091
  "priority": 2,
@@ -2222,7 +2222,7 @@
2222
  ],
2223
  "status": "planning_artifact"
2224
  },
2225
- "generated_at_utc": "2026-06-06T23:26:13+00:00",
2226
  "omni_plan": {
2227
  "adapter": "LoRA rank 16, alpha 32, dropout 0.05",
2228
  "backbone": "Qwen/Qwen3-Omni-30B-A3B-Instruct",
@@ -2332,6 +2332,27 @@
2332
  "stage": "future",
2333
  "status": "verified_companion_result"
2334
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2335
  {
2336
  "completion_evidence": [
2337
  "error-analysis tables",
@@ -2362,6 +2383,7 @@
2362
  "deliverables": [
2363
  "backbone registry",
2364
  "Cosmos 3 world-model branch plan",
 
2365
  "Qwen3-Omni LoRA baseline plan",
2366
  "OpenVLA/openpi/GR00T policy-branch candidates",
2367
  "model-specific evaluation additions"
@@ -2369,9 +2391,9 @@
2369
  "entry_condition": "The selected episodes are prepared or a 3-8 episode dry run is available for preprocessing checks.",
2370
  "id": "foundation_model_selection_matrix",
2371
  "name": "Foundation-Model Selection Matrix",
2372
- "reader_takeaway": "Qwen3-Omni remains the first trainable held-out pilot; Cosmos 3 is the first world-model branch. Cosmos3-Super now has camera-pose proxy forward-dynamics targets ready for trainer implementation, while VLA/policy models wait for robot-compatible action targets.",
2373
- "stage": "omni",
2374
- "status": "next"
2375
  },
2376
  {
2377
  "completion_evidence": [
@@ -2392,16 +2414,17 @@
2392
  "name": "64-128 Episode Robustness Run",
2393
  "reader_takeaway": "The robustness run tests whether the pilot conclusions survive broader sessions and missing modalities.",
2394
  "stage": "future",
2395
- "status": "planned"
2396
  },
2397
  {
2398
  "completion_evidence": [
2399
  "task-specific held-out evaluations",
 
2400
  "qualitative inspection",
2401
  "updated model cards"
2402
  ],
2403
  "deliverables": [
2404
- "Cosmos 3 future-window or action-conditioned world-model probe",
2405
  "OpenVLA/openpi/GR00T action-policy baseline",
2406
  "audio/video/depth/pose/mocap conditioning checks",
2407
  "affordance and object-interaction tasks",
@@ -2410,7 +2433,7 @@
2410
  "entry_condition": "Enough multi-episode data, compute budget, and model-specific action/world-state targets.",
2411
  "id": "foundation_world_model_extensions",
2412
  "name": "Cosmos 3 and Policy-Model Extensions",
2413
- "reader_takeaway": "The long-term direction is richer multimodal representation learning for embodied-AI reasoning, with model branches chosen by task fit rather than by a single default backbone.",
2414
  "stage": "future",
2415
  "status": "planned"
2416
  },
 
2040
  "step": 2
2041
  },
2042
  {
2043
+ "action": "Run 3-8 episode dry runs for any next backbone before scaling beyond the selected split.",
2044
  "name": "Model-selection dry run",
2045
  "step": 3
2046
  },
2047
  {
2048
+ "action": "Promote Cosmos 3 beyond the current Nano compatibility and Super forward-dynamics runs only when loss metrics, preprocessing, and storage justify the added compute.",
2049
  "name": "World-model branch",
2050
  "step": 4
2051
  },
 
2084
  {
2085
  "best_role": "Embodied world modeling, action generation, future-window prediction, and synthetic-data expansion.",
2086
  "category": "world_foundation_model",
2087
+ "current_decision": "implemented_as_nano_future_window_and_super_forward_dynamics_branches",
2088
+ "entry_condition": "Use separate metrics for Nano future-window retrieval and Super forward-dynamics MSE; do not compare them directly to Qwen JSON-task accuracy.",
2089
  "family": "Cosmos 3",
2090
  "openness": "track_official_nvidia_release_and_available_weights",
2091
  "priority": 2,
 
2222
  ],
2223
  "status": "planning_artifact"
2224
  },
2225
+ "generated_at_utc": "2026-06-08T12:22:13+00:00",
2226
  "omni_plan": {
2227
  "adapter": "LoRA rank 16, alpha 32, dropout 0.05",
2228
  "backbone": "Qwen/Qwen3-Omni-30B-A3B-Instruct",
 
2332
  "stage": "future",
2333
  "status": "verified_companion_result"
2334
  },
2335
+ {
2336
+ "completion_evidence": [
2337
+ "TASK_SUITE_ENHANCEMENT_128.md",
2338
+ "docs/data/task_suite_enhancement_128.json",
2339
+ "results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json",
2340
+ "scripts/omni/build_task_suite_enhancement_128.py"
2341
+ ],
2342
+ "deliverables": [
2343
+ "dense-window and multiscale export estimates",
2344
+ "hierarchical action/subtask target contract",
2345
+ "raw-feature shard priorities for unsupported tasks",
2346
+ "Qwen v5 and Cosmos continuation run cards",
2347
+ "publication-ready enhancement artifacts"
2348
+ ],
2349
+ "entry_condition": "Same selected 96/16/16 split and current public 3,808-window export.",
2350
+ "id": "task_suite_enhancement_128",
2351
+ "name": "128-Episode Task Suite Enhancement Pack",
2352
+ "reader_takeaway": "The current 128-episode setup still has headroom: use multiscale_20s10_40s20_80s40, hierarchical labels, label-normalized scoring, and raw-feature shards before adding more episodes.",
2353
+ "stage": "future",
2354
+ "status": "current"
2355
+ },
2356
  {
2357
  "completion_evidence": [
2358
  "error-analysis tables",
 
2383
  "deliverables": [
2384
  "backbone registry",
2385
  "Cosmos 3 world-model branch plan",
2386
+ "Cosmos3-Super Forward-Dynamics LoRA verified package",
2387
  "Qwen3-Omni LoRA baseline plan",
2388
  "OpenVLA/openpi/GR00T policy-branch candidates",
2389
  "model-specific evaluation additions"
 
2391
  "entry_condition": "The selected episodes are prepared or a 3-8 episode dry run is available for preprocessing checks.",
2392
  "id": "foundation_model_selection_matrix",
2393
  "name": "Foundation-Model Selection Matrix",
2394
+ "reader_takeaway": "Qwen3-Omni remains the structured JSON held-out pilot; Cosmos 3 is the first world-model branch. Cosmos3-Super now has a verified forward-dynamics LoRA over camera-pose proxy targets, while VLA/policy models wait for robot-compatible action targets.",
2395
+ "stage": "future",
2396
+ "status": "current"
2397
  },
2398
  {
2399
  "completion_evidence": [
 
2414
  "name": "64-128 Episode Robustness Run",
2415
  "reader_takeaway": "The robustness run tests whether the pilot conclusions survive broader sessions and missing modalities.",
2416
  "stage": "future",
2417
+ "status": "partially_implemented"
2418
  },
2419
  {
2420
  "completion_evidence": [
2421
  "task-specific held-out evaluations",
2422
+ "verified Cosmos3-Super forward-dynamics LoRA package",
2423
  "qualitative inspection",
2424
  "updated model cards"
2425
  ],
2426
  "deliverables": [
2427
+ "Cosmos 3 future-window and action-conditioned world-model probes",
2428
  "OpenVLA/openpi/GR00T action-policy baseline",
2429
  "audio/video/depth/pose/mocap conditioning checks",
2430
  "affordance and object-interaction tasks",
 
2433
  "entry_condition": "Enough multi-episode data, compute budget, and model-specific action/world-state targets.",
2434
  "id": "foundation_world_model_extensions",
2435
  "name": "Cosmos 3 and Policy-Model Extensions",
2436
+ "reader_takeaway": "The Cosmos branch now includes Nano future-window compatibility and Super forward-dynamics LoRA; the long-term direction remains richer multimodal representation learning with model branches chosen by task fit rather than by a single default backbone.",
2437
  "stage": "future",
2438
  "status": "planned"
2439
  },
data/scope_claims_audit.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "status": "pass",
3
- "generated_at_utc": "2026-06-07T15:47:31+00:00",
4
  "summary": {
5
  "qwen3_omni_verified_diagnostic_pilot": true,
6
  "dataset_manifest_num_episodes": 119,
@@ -9,7 +9,7 @@
9
  "eval_num_samples": 448,
10
  "eval_json_validity_rate": 1.0,
11
  "quality_target_met": true,
12
- "historical_identifier_count": 1545,
13
  "public_32_episode_status_file_count": 1,
14
  "failure_count": 0
15
  },
@@ -25,7 +25,7 @@
25
  {
26
  "name": "summary_metrics_preserves_verified_diagnostic_status",
27
  "status": "pass",
28
- "detail": "The selected-episode Qwen3-Omni diagnostic pilot is verified on the 96/16/16 split and now meets the 98% target for JSON validity; action/subtask quality remains weak, so current results are diagnostic baselines, not strong model-quality claims.",
29
  "evidence": [
30
  "docs/data/summary_metrics.json"
31
  ]
@@ -35,7 +35,7 @@
35
  "status": "pass",
36
  "detail": "episodes=119, samples=3808, split_counts={'train': 2848, 'val': 512, 'test': 448}",
37
  "evidence": [
38
- "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v3_strict_label_prompt_reuse_lora_eval_test_full/dataset/dataset_manifest.json"
39
  ]
40
  },
41
  {
@@ -43,7 +43,7 @@
43
  "status": "pass",
44
  "detail": "train=2848, val=512, processes=8",
45
  "evidence": [
46
- "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v3_strict_label_prompt_reuse_lora_eval_test_full/training/training_metadata.json"
47
  ]
48
  },
49
  {
@@ -51,7 +51,7 @@
51
  "status": "pass",
52
  "detail": "samples=448, split=test, held_out=14, json_validity=1.0",
53
  "evidence": [
54
- "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v3_strict_label_prompt_reuse_lora_eval_test_full/eval/metrics.json"
55
  ]
56
  },
57
  {
@@ -59,7 +59,7 @@
59
  "status": "pass",
60
  "detail": "audit_status=pass, issues=0",
61
  "evidence": [
62
- "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v3_strict_label_prompt_reuse_lora_eval_test_full/package_audit.json"
63
  ]
64
  },
65
  {
@@ -84,7 +84,7 @@
84
  {
85
  "name": "historical_32ep_identifiers_are_confined_to_readiness_artifacts",
86
  "status": "pass",
87
- "detail": "historical identifiers found in result provenance files=1545",
88
  "evidence": [
89
  "results/omni_finetune/"
90
  ]
@@ -424,6 +424,6 @@
424
  "example": "{\"id\": \"xperience-10m-sample:qa:53\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 1060, \"end_frame\": 1079, \"num_frames\": 20}, \"media\": {\"video_path"
425
  }
426
  ],
427
- "historical_identifier_total_count": 1545,
428
  "failures": []
429
  }
 
1
  {
2
  "status": "pass",
3
+ "generated_at_utc": "2026-06-11T07:10:47+00:00",
4
  "summary": {
5
  "qwen3_omni_verified_diagnostic_pilot": true,
6
  "dataset_manifest_num_episodes": 119,
 
9
  "eval_num_samples": 448,
10
  "eval_json_validity_rate": 1.0,
11
  "quality_target_met": true,
12
+ "historical_identifier_count": 1799,
13
  "public_32_episode_status_file_count": 1,
14
  "failure_count": 0
15
  },
 
25
  {
26
  "name": "summary_metrics_preserves_verified_diagnostic_status",
27
  "status": "pass",
28
+ "detail": "The selected-episode Qwen3-Omni diagnostic pilot is verified on the 96/16/16 split and meets the 98% target for JSON validity; action/subtask quality remains weak, so it is a structured-task baseline. Cosmos3-Nano future-window compatibility and Cosmos3-Super Forward-Dynamics LoRA are also verified as separate world-model diagnostics with different metrics.",
29
  "evidence": [
30
  "docs/data/summary_metrics.json"
31
  ]
 
35
  "status": "pass",
36
  "detail": "episodes=119, samples=3808, split_counts={'train': 2848, 'val': 512, 'test': 448}",
37
  "evidence": [
38
+ "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/dataset/dataset_manifest.json"
39
  ]
40
  },
41
  {
 
43
  "status": "pass",
44
  "detail": "train=2848, val=512, processes=8",
45
  "evidence": [
46
+ "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/training/training_metadata.json"
47
  ]
48
  },
49
  {
 
51
  "status": "pass",
52
  "detail": "samples=448, split=test, held_out=14, json_validity=1.0",
53
  "evidence": [
54
+ "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/eval/metrics.json"
55
  ]
56
  },
57
  {
 
59
  "status": "pass",
60
  "detail": "audit_status=pass, issues=0",
61
  "evidence": [
62
+ "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full/package_audit.json"
63
  ]
64
  },
65
  {
 
84
  {
85
  "name": "historical_32ep_identifiers_are_confined_to_readiness_artifacts",
86
  "status": "pass",
87
+ "detail": "historical identifiers found in result provenance files=1799",
88
  "evidence": [
89
  "results/omni_finetune/"
90
  ]
 
424
  "example": "{\"id\": \"xperience-10m-sample:qa:53\", \"episode_id\": \"xperience-10m-sample\", \"split\": \"train\", \"target\": \"episode_qa\", \"prompt_type\": \"json_episode_understanding\", \"center_window\": {\"start_frame\": 1060, \"end_frame\": 1079, \"num_frames\": 20}, \"media\": {\"video_path"
425
  }
426
  ],
427
+ "historical_identifier_total_count": 1799,
428
  "failures": []
429
  }
data/source_alignment_audit.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "title": "Ropedia Xperience-10M Source Alignment Note",
3
  "status": "pass",
4
- "generated_at_utc": "2026-06-04T16:48:58+00:00",
5
  "alignment_json": "docs/data/xperience10m_dataset_card_alignment.json",
6
  "alignment_summary": {
7
  "full_dataset_repo": "ropedia-ai/xperience-10m",
 
1
  {
2
  "title": "Ropedia Xperience-10M Source Alignment Note",
3
  "status": "pass",
4
+ "generated_at_utc": "2026-06-11T07:10:46+00:00",
5
  "alignment_json": "docs/data/xperience10m_dataset_card_alignment.json",
6
  "alignment_summary": {
7
  "full_dataset_repo": "ropedia-ai/xperience-10m",
data/summary_metrics.json CHANGED
@@ -14,7 +14,7 @@
14
  "visualization.rrd"
15
  ],
16
  "access_status": "The gated Xperience-10M dataset is available for selected multi-episode pilot preparation.",
17
- "current_scope": "The selected-episode Qwen3-Omni diagnostic pilot is verified on the 96/16/16 split and now meets the 98% target for JSON validity; action/subtask quality remains weak, so current results are diagnostic baselines, not strong model-quality claims."
18
  },
19
  "models": {
20
  "motion_action": {
 
14
  "visualization.rrd"
15
  ],
16
  "access_status": "The gated Xperience-10M dataset is available for selected multi-episode pilot preparation.",
17
+ "current_scope": "The selected-episode Qwen3-Omni diagnostic pilot is verified on the 96/16/16 split and meets the 98% target for JSON validity; action/subtask quality remains weak, so it is a structured-task baseline. Cosmos3-Nano future-window compatibility and Cosmos3-Super Forward-Dynamics LoRA are also verified as separate world-model diagnostics with different metrics."
18
  },
19
  "models": {
20
  "motion_action": {
data/task_suite_enhancement_128.json ADDED
@@ -0,0 +1,599 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "title": "Ropedia Xperience-10M 128-Episode Task Suite Enhancement Pack",
3
+ "status": "pass",
4
+ "run_id": "task_suite_enhancement_128_v1_20260608",
5
+ "generated_at_utc": "2026-06-08T12:30:01+00:00",
6
+ "scope": "No-new-episode enhancement plan over the current selected 128-episode 96/16/16 split.",
7
+ "current_128_split": {
8
+ "total_windows": 3808,
9
+ "split_windows": {
10
+ "test": 448,
11
+ "train": 2848,
12
+ "val": 512
13
+ },
14
+ "selected_episode_counts": {
15
+ "test": 16,
16
+ "train": 96,
17
+ "val": 16
18
+ },
19
+ "windowed_episode_counts": {
20
+ "test": 14,
21
+ "train": 89,
22
+ "val": 16
23
+ },
24
+ "unique_main_tasks": 106,
25
+ "windows_per_episode": {
26
+ "min": 32,
27
+ "median": 32,
28
+ "max": 32
29
+ }
30
+ },
31
+ "dense_window_scenarios": [
32
+ {
33
+ "id": "current_export",
34
+ "window_frames": 20,
35
+ "stride_frames": "selected_sparse_windows",
36
+ "role": "current public 128-episode JSON-task export",
37
+ "estimated_windows": 3808,
38
+ "estimated_split_windows": {
39
+ "test": 448,
40
+ "train": 2848,
41
+ "val": 512
42
+ },
43
+ "multiplier_vs_current_export": 1.0,
44
+ "source_note": "Estimated from current public-safe window frame spans; the real exporter must still validate raw-stream availability and label coverage."
45
+ },
46
+ {
47
+ "id": "dense_20f_stride20",
48
+ "window_frames": 20,
49
+ "stride_frames": 20,
50
+ "role": "non-overlap dense coverage over each observed episode frame span",
51
+ "estimated_windows": 30422,
52
+ "estimated_split_windows": {
53
+ "test": 3383,
54
+ "train": 22822,
55
+ "val": 4217
56
+ },
57
+ "multiplier_vs_current_export": 7.99,
58
+ "source_note": "Estimated from current public-safe window frame spans; the real exporter must still validate raw-stream availability and label coverage."
59
+ },
60
+ {
61
+ "id": "dense_20f_stride10",
62
+ "window_frames": 20,
63
+ "stride_frames": 10,
64
+ "role": "2x overlap action/subtask densification",
65
+ "estimated_windows": 60725,
66
+ "estimated_split_windows": {
67
+ "test": 6752,
68
+ "train": 45555,
69
+ "val": 8418
70
+ },
71
+ "multiplier_vs_current_export": 15.95,
72
+ "source_note": "Estimated from current public-safe window frame spans; the real exporter must still validate raw-stream availability and label coverage."
73
+ },
74
+ {
75
+ "id": "dense_20f_stride5",
76
+ "window_frames": 20,
77
+ "stride_frames": 5,
78
+ "role": "high-overlap action boundary and transition stress setting",
79
+ "estimated_windows": 121331,
80
+ "estimated_split_windows": {
81
+ "test": 13490,
82
+ "train": 91021,
83
+ "val": 16820
84
+ },
85
+ "multiplier_vs_current_export": 31.86,
86
+ "source_note": "Estimated from current public-safe window frame spans; the real exporter must still validate raw-stream availability and label coverage."
87
+ },
88
+ {
89
+ "id": "medium_40f_stride20",
90
+ "window_frames": 40,
91
+ "stride_frames": 20,
92
+ "role": "subtask/procedure context window",
93
+ "estimated_windows": 30303,
94
+ "estimated_split_windows": {
95
+ "test": 3369,
96
+ "train": 22733,
97
+ "val": 4201
98
+ },
99
+ "multiplier_vs_current_export": 7.96,
100
+ "source_note": "Estimated from current public-safe window frame spans; the real exporter must still validate raw-stream availability and label coverage."
101
+ },
102
+ {
103
+ "id": "long_80f_stride40",
104
+ "window_frames": 80,
105
+ "stride_frames": 40,
106
+ "role": "procedure and world-model context window",
107
+ "estimated_windows": 15067,
108
+ "estimated_split_windows": {
109
+ "test": 1674,
110
+ "train": 11305,
111
+ "val": 2088
112
+ },
113
+ "multiplier_vs_current_export": 3.96,
114
+ "source_note": "Estimated from current public-safe window frame spans; the real exporter must still validate raw-stream availability and label coverage."
115
+ },
116
+ {
117
+ "id": "multiscale_20s10_40s20_80s40",
118
+ "role": "recommended no-new-episode v5 export: short action windows plus medium/long procedure context",
119
+ "components": [
120
+ "dense_20f_stride10",
121
+ "medium_40f_stride20",
122
+ "long_80f_stride40"
123
+ ],
124
+ "estimated_windows": 106095,
125
+ "estimated_split_windows": {
126
+ "test": 11795,
127
+ "train": 79593,
128
+ "val": 14707
129
+ },
130
+ "multiplier_vs_current_export": 27.86,
131
+ "source_note": "Composite planning estimate; store as a new export run rather than replacing existing 128-episode packages."
132
+ }
133
+ ],
134
+ "hierarchical_target_contract": {
135
+ "id": "xperience10m_128_hierarchical_action_targets_v1",
136
+ "status": "ready_for_export",
137
+ "purpose": "Reduce fine-grained label sparsity without changing the sealed 96/16/16 episode split.",
138
+ "target_fields": [
139
+ {
140
+ "field": "action_family",
141
+ "source": "normalized true action string",
142
+ "values": [
143
+ "locomotion",
144
+ "reach_grasp_release",
145
+ "place_arrange_align",
146
+ "manipulate_adjust",
147
+ "tool_cut_mark_write",
148
+ "sort_count_organize",
149
+ "inspect_observe_use",
150
+ "clean_cook",
151
+ "other_fine_action",
152
+ "unknown"
153
+ ],
154
+ "metric": "macro_f1"
155
+ },
156
+ {
157
+ "field": "action_verb",
158
+ "source": "first normalized verb phrase from action label",
159
+ "metric": "macro_f1 with train-seen and unseen slices"
160
+ },
161
+ {
162
+ "field": "fine_action",
163
+ "source": "existing action label",
164
+ "metric": "exact match and label-normalized semantic family match"
165
+ },
166
+ {
167
+ "field": "subtask_family",
168
+ "source": "normalized subtask phrase or main task fallback",
169
+ "metric": "accuracy and macro_f1"
170
+ },
171
+ {
172
+ "field": "contact_transition",
173
+ "source": "existing contact and transition fields",
174
+ "metric": "accuracy, balanced accuracy, calibration"
175
+ },
176
+ {
177
+ "field": "object_set",
178
+ "source": "existing objects list",
179
+ "metric": "micro_f1 and object-category recall"
180
+ }
181
+ ],
182
+ "public_safety": [
183
+ "No raw MP4/HDF5/RRD files are written.",
184
+ "No full Qwen/Cosmos weights are mirrored.",
185
+ "Generated labels and aggregate metrics remain public-safe derived metadata."
186
+ ]
187
+ },
188
+ "task_bottlenecks": [
189
+ {
190
+ "task": "next_action",
191
+ "display_name": "Next-Action Prediction",
192
+ "priority": "highest",
193
+ "simple_status": "pass",
194
+ "simple_primary_metric": "macro_f1",
195
+ "simple_primary_score": 0.00019966057701906761,
196
+ "neural_status": "pass",
197
+ "neural_primary_score": 0.0,
198
+ "num_classes": 1184,
199
+ "unseen_test_class_count": 145,
200
+ "bottleneck": "fine-grained label explosion and held-out unseen labels",
201
+ "next_action": "add hierarchical action/subtask families plus label-normalized scoring"
202
+ },
203
+ {
204
+ "task": "timeline_action",
205
+ "display_name": "Action Recognition",
206
+ "priority": "highest",
207
+ "simple_status": "pass",
208
+ "simple_primary_metric": "macro_f1",
209
+ "simple_primary_score": 0.00017511601435951318,
210
+ "neural_status": "pass",
211
+ "neural_primary_score": 0.0,
212
+ "num_classes": 1187,
213
+ "unseen_test_class_count": 144,
214
+ "bottleneck": "fine-grained label explosion and held-out unseen labels",
215
+ "next_action": "add hierarchical action/subtask families plus label-normalized scoring"
216
+ },
217
+ {
218
+ "task": "timeline_subtask",
219
+ "display_name": "Procedure Step Recognition",
220
+ "priority": "highest",
221
+ "simple_status": "pass",
222
+ "simple_primary_metric": "macro_f1",
223
+ "simple_primary_score": 0.0,
224
+ "neural_status": "pass",
225
+ "neural_primary_score": 0.0,
226
+ "num_classes": 850,
227
+ "unseen_test_class_count": 113,
228
+ "bottleneck": "fine-grained label explosion and held-out unseen labels",
229
+ "next_action": "add hierarchical action/subtask families plus label-normalized scoring"
230
+ },
231
+ {
232
+ "task": "cross_modal_retrieval",
233
+ "display_name": "Cross-Modal Retrieval",
234
+ "priority": "high",
235
+ "simple_status": "unsupported_without_raw_128_feature_blocks",
236
+ "simple_primary_metric": "mrr",
237
+ "simple_primary_score": null,
238
+ "neural_status": "not_run",
239
+ "neural_primary_score": null,
240
+ "num_classes": null,
241
+ "unseen_test_class_count": null,
242
+ "bottleneck": "missing raw 128-episode feature blocks",
243
+ "next_action": "export compact raw-feature shards for this task before model comparison"
244
+ },
245
+ {
246
+ "task": "hand_trajectory_forecast",
247
+ "display_name": "Hand Trajectory Forecasting",
248
+ "priority": "high",
249
+ "simple_status": "unsupported_without_raw_128_feature_blocks",
250
+ "simple_primary_metric": "mpjpe",
251
+ "simple_primary_score": null,
252
+ "neural_status": "not_run",
253
+ "neural_primary_score": null,
254
+ "num_classes": null,
255
+ "unseen_test_class_count": null,
256
+ "bottleneck": "missing raw 128-episode feature blocks",
257
+ "next_action": "export compact raw-feature shards for this task before model comparison"
258
+ },
259
+ {
260
+ "task": "misalignment_detection",
261
+ "display_name": "Multimodal Synchronization Detection",
262
+ "priority": "high",
263
+ "simple_status": "unsupported_without_raw_128_feature_blocks",
264
+ "simple_primary_metric": "f1",
265
+ "simple_primary_score": null,
266
+ "neural_status": "not_run",
267
+ "neural_primary_score": null,
268
+ "num_classes": null,
269
+ "unseen_test_class_count": null,
270
+ "bottleneck": "missing raw 128-episode feature blocks",
271
+ "next_action": "export compact raw-feature shards for this task before model comparison"
272
+ },
273
+ {
274
+ "task": "modality_reconstruction",
275
+ "display_name": "Cross-Modal Reconstruction",
276
+ "priority": "high",
277
+ "simple_status": "unsupported_without_raw_128_feature_blocks",
278
+ "simple_primary_metric": "r2",
279
+ "simple_primary_score": null,
280
+ "neural_status": "not_run",
281
+ "neural_primary_score": null,
282
+ "num_classes": null,
283
+ "unseen_test_class_count": null,
284
+ "bottleneck": "missing raw 128-episode feature blocks",
285
+ "next_action": "export compact raw-feature shards for this task before model comparison"
286
+ },
287
+ {
288
+ "task": "caption_grounding",
289
+ "display_name": "Language Grounding",
290
+ "priority": "medium",
291
+ "simple_status": "pass",
292
+ "simple_primary_metric": "mrr",
293
+ "simple_primary_score": 0.012785504572093487,
294
+ "neural_status": "not_run",
295
+ "neural_primary_score": null,
296
+ "num_classes": null,
297
+ "unseen_test_class_count": null,
298
+ "bottleneck": "weak public-safe metadata/text baseline",
299
+ "next_action": "add dense windows and stronger fusion baselines before interpreting model quality"
300
+ },
301
+ {
302
+ "task": "contact_prediction",
303
+ "display_name": "Contact State Prediction",
304
+ "priority": "medium",
305
+ "simple_status": "pass",
306
+ "simple_primary_metric": "macro_f1",
307
+ "simple_primary_score": 0.5167950693374422,
308
+ "neural_status": "pass",
309
+ "neural_primary_score": 0.21951219512195122,
310
+ "num_classes": 2,
311
+ "unseen_test_class_count": 0,
312
+ "bottleneck": "usable control task",
313
+ "next_action": "keep as sanity/control metric for future dense-window and model runs"
314
+ },
315
+ {
316
+ "task": "object_relevance",
317
+ "display_name": "Object Relevance Prediction",
318
+ "priority": "medium",
319
+ "simple_status": "pass",
320
+ "simple_primary_metric": "micro_f1",
321
+ "simple_primary_score": 0.18221614227086183,
322
+ "neural_status": "pass",
323
+ "neural_primary_score": 0.1053878034339846,
324
+ "num_classes": null,
325
+ "unseen_test_class_count": null,
326
+ "bottleneck": "moderate task signal, still needs robustness split",
327
+ "next_action": "add session/task-family slices and bootstrap confidence intervals"
328
+ },
329
+ {
330
+ "task": "temporal_order",
331
+ "display_name": "Temporal Order Verification",
332
+ "priority": "medium",
333
+ "simple_status": "pass",
334
+ "simple_primary_metric": "f1",
335
+ "simple_primary_score": 0.32713178294573647,
336
+ "neural_status": "not_run",
337
+ "neural_primary_score": null,
338
+ "num_classes": 2,
339
+ "unseen_test_class_count": null,
340
+ "bottleneck": "usable control task",
341
+ "next_action": "keep as sanity/control metric for future dense-window and model runs"
342
+ },
343
+ {
344
+ "task": "transition_detection",
345
+ "display_name": "Action Boundary Detection",
346
+ "priority": "medium",
347
+ "simple_status": "pass",
348
+ "simple_primary_metric": "macro_f1",
349
+ "simple_primary_score": 0.5219803670507895,
350
+ "neural_status": "pass",
351
+ "neural_primary_score": 0.45822172492907925,
352
+ "num_classes": 2,
353
+ "unseen_test_class_count": 0,
354
+ "bottleneck": "usable control task",
355
+ "next_action": "keep as sanity/control metric for future dense-window and model runs"
356
+ }
357
+ ],
358
+ "qwen_v4_error_pressure": {
359
+ "run_id": "xperience10m_qwen3_omni_128ep_structured_json_v4_4epoch_full8gpu_lora_eval_test_full",
360
+ "samples": 448,
361
+ "json_validity_rate": 1.0,
362
+ "action_macro_f1": 0.0018678269676001454,
363
+ "subtask_accuracy": 0.0,
364
+ "next_action_accuracy": 0.033482142857142856,
365
+ "contact_accuracy": 0.7299107142857143,
366
+ "transition_accuracy": 0.9732142857142857,
367
+ "object_micro_f1": 0.31099781500364165,
368
+ "num_unseen_label_samples": 317,
369
+ "unseen_label_sample_share": 0.7075892857142857,
370
+ "seen_label_accuracy": 0.09923664122137404,
371
+ "unseen_label_accuracy": 0.0031545741324921135,
372
+ "eval_unique_labels": 189,
373
+ "eval_singleton_label_count": 42,
374
+ "eval_singleton_label_share": 0.2222222222222222,
375
+ "action_family_error_summary": [
376
+ {
377
+ "family": "manipulate_adjust",
378
+ "samples": 98,
379
+ "action_exact_rate": 0.030612244897959183,
380
+ "seen_share": 0.22448979591836735,
381
+ "contact_exact_rate": 0.7959183673469388,
382
+ "transition_exact_rate": 1.0
383
+ },
384
+ {
385
+ "family": "reach_grasp_release",
386
+ "samples": 88,
387
+ "action_exact_rate": 0.011363636363636364,
388
+ "seen_share": 0.45454545454545453,
389
+ "contact_exact_rate": 0.7954545454545454,
390
+ "transition_exact_rate": 0.9318181818181818
391
+ },
392
+ {
393
+ "family": "other_fine_action",
394
+ "samples": 73,
395
+ "action_exact_rate": 0.0,
396
+ "seen_share": 0.2465753424657534,
397
+ "contact_exact_rate": 0.7945205479452054,
398
+ "transition_exact_rate": 0.9726027397260274
399
+ },
400
+ {
401
+ "family": "place_arrange_align",
402
+ "samples": 65,
403
+ "action_exact_rate": 0.03076923076923077,
404
+ "seen_share": 0.26153846153846155,
405
+ "contact_exact_rate": 0.5384615384615384,
406
+ "transition_exact_rate": 0.9692307692307692
407
+ },
408
+ {
409
+ "family": "sort_count_organize",
410
+ "samples": 36,
411
+ "action_exact_rate": 0.0,
412
+ "seen_share": 0.1388888888888889,
413
+ "contact_exact_rate": 0.6388888888888888,
414
+ "transition_exact_rate": 1.0
415
+ },
416
+ {
417
+ "family": "tool_cut_mark_write",
418
+ "samples": 28,
419
+ "action_exact_rate": 0.25,
420
+ "seen_share": 0.6428571428571429,
421
+ "contact_exact_rate": 1.0,
422
+ "transition_exact_rate": 1.0
423
+ },
424
+ {
425
+ "family": "inspect_observe_use",
426
+ "samples": 27,
427
+ "action_exact_rate": 0.0,
428
+ "seen_share": 0.37037037037037035,
429
+ "contact_exact_rate": 0.6666666666666666,
430
+ "transition_exact_rate": 0.9629629629629629
431
+ },
432
+ {
433
+ "family": "locomotion",
434
+ "samples": 27,
435
+ "action_exact_rate": 0.0,
436
+ "seen_share": 0.037037037037037035,
437
+ "contact_exact_rate": 0.48148148148148145,
438
+ "transition_exact_rate": 1.0
439
+ },
440
+ {
441
+ "family": "clean_cook",
442
+ "samples": 6,
443
+ "action_exact_rate": 0.16666666666666666,
444
+ "seen_share": 0.0,
445
+ "contact_exact_rate": 0.6666666666666666,
446
+ "transition_exact_rate": 0.8333333333333334
447
+ }
448
+ ],
449
+ "top_true_objects": [
450
+ {
451
+ "object": "smartphone",
452
+ "count": 134
453
+ },
454
+ {
455
+ "object": "table",
456
+ "count": 56
457
+ },
458
+ {
459
+ "object": "scissors",
460
+ "count": 47
461
+ },
462
+ {
463
+ "object": "water bottle",
464
+ "count": 43
465
+ },
466
+ {
467
+ "object": "pen",
468
+ "count": 41
469
+ },
470
+ {
471
+ "object": "paper",
472
+ "count": 34
473
+ },
474
+ {
475
+ "object": "cardboard",
476
+ "count": 32
477
+ },
478
+ {
479
+ "object": "utility knife",
480
+ "count": 32
481
+ },
482
+ {
483
+ "object": "marker",
484
+ "count": 31
485
+ },
486
+ {
487
+ "object": "puzzle box",
488
+ "count": 31
489
+ },
490
+ {
491
+ "object": "paper strips",
492
+ "count": 29
493
+ },
494
+ {
495
+ "object": "buttons",
496
+ "count": 28
497
+ },
498
+ {
499
+ "object": "cardboard box",
500
+ "count": 25
501
+ },
502
+ {
503
+ "object": "ruler",
504
+ "count": 23
505
+ },
506
+ {
507
+ "object": "power bank",
508
+ "count": 21
509
+ },
510
+ {
511
+ "object": "star beads",
512
+ "count": 21
513
+ },
514
+ {
515
+ "object": "cardboard pieces",
516
+ "count": 20
517
+ },
518
+ {
519
+ "object": "hand",
520
+ "count": 19
521
+ },
522
+ {
523
+ "object": "canned food",
524
+ "count": 19
525
+ },
526
+ {
527
+ "object": "jigsaw puzzle",
528
+ "count": 19
529
+ }
530
+ ]
531
+ },
532
+ "cosmos3_super_forward_dynamics_reference": {
533
+ "status": "verified",
534
+ "run_id": null,
535
+ "train_rows": null,
536
+ "val_rows": null,
537
+ "test_rows": null,
538
+ "test_mse": null,
539
+ "adapter_parameter_numel": null
540
+ },
541
+ "experiment_backlog": [
542
+ {
543
+ "id": "dense_window_export_v1",
544
+ "priority": 1,
545
+ "status": "ready_to_implement",
546
+ "goal": "Create a new dense-window export over the same 128 episodes without replacing existing JSONL packages.",
547
+ "expected_artifacts": [
548
+ "dataset_dense_20f_stride10.jsonl",
549
+ "dataset_dense_multiscale_manifest.json",
550
+ "label_family_distribution.json"
551
+ ],
552
+ "gate": "episode ids and split assignment must exactly match the current 96/16/16 split"
553
+ },
554
+ {
555
+ "id": "hierarchical_qwen3_v5",
556
+ "priority": 2,
557
+ "status": "ready_after_dense_export",
558
+ "goal": "Train/evaluate Qwen3 with hierarchical action/subtask targets, constrained label options, and no-public-overwrite packaging.",
559
+ "suggested_setup": "high-rank LoRA or partial projector/last-layer unfreeze before full-parameter tuning",
560
+ "primary_comparison": "Qwen3 v4 action/subtask/next-action plus seen/unseen-label slices"
561
+ },
562
+ {
563
+ "id": "raw_feature_unblocker_128",
564
+ "priority": 3,
565
+ "status": "ready_to_implement_on_training_host",
566
+ "goal": "Export compact 128-episode raw feature shards for tasks currently marked unsupported_without_raw_128_feature_blocks.",
567
+ "target_tasks": [
568
+ "hand_trajectory_forecast",
569
+ "cross_modal_retrieval",
570
+ "modality_reconstruction",
571
+ "misalignment_detection"
572
+ ]
573
+ },
574
+ {
575
+ "id": "cosmos3_fd_v2_multiscale",
576
+ "priority": 4,
577
+ "status": "ready_after_dense_export",
578
+ "goal": "Continue Cosmos3-Super forward-dynamics with multiscale horizons and temporal consistency metrics.",
579
+ "primary_comparison": "Cosmos3-Super Forward-Dynamics v1 validation/test MSE and rank-level loss records"
580
+ },
581
+ {
582
+ "id": "robustness_and_confidence_pack",
583
+ "priority": 5,
584
+ "status": "ready_from_existing_outputs",
585
+ "goal": "Add bootstrap confidence intervals, task-family slices, session slices, and random-time/random-label sanity checks.",
586
+ "public_output": "results/omni_finetune/task_suite_enhancement_128_v1_20260608/robustness_pack_v1.json"
587
+ }
588
+ ],
589
+ "public_artifacts": {
590
+ "result_dir": "results/omni_finetune/task_suite_enhancement_128_v1_20260608",
591
+ "public_json": "docs/data/task_suite_enhancement_128.json",
592
+ "public_markdown": "TASK_SUITE_ENHANCEMENT_128.md"
593
+ },
594
+ "non_overwrite_policy": {
595
+ "result_directory_created_once": true,
596
+ "stable_public_summaries_update_to_latest_enhancement_pack": true,
597
+ "prior_model_result_packages_overwritten": false
598
+ }
599
+ }
data/task_surface_integrity.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "status": "pass",
3
- "generated_at_utc": "2026-06-07T15:47:30+00:00",
4
  "summary": {
5
  "task_count": 12,
6
  "expected_task_count": 12,
@@ -64,33 +64,33 @@
64
  "observed": "timeline_action"
65
  },
66
  {
67
- "name": "timeline_action: public_field_input_short_is_human_readable",
68
  "status": "pass",
69
- "value": "20-frame multimodal window",
70
  "raw_hits": []
71
  },
72
  {
73
- "name": "timeline_action: public_field_research_name_is_human_readable",
74
  "status": "pass",
75
- "value": "Egocentric Action Recognition",
76
  "raw_hits": []
77
  },
78
  {
79
- "name": "timeline_action: public_field_display_name_is_human_readable",
80
  "status": "pass",
81
- "value": "Action Recognition",
82
  "raw_hits": []
83
  },
84
  {
85
- "name": "timeline_action: public_field_card_blurb_is_human_readable",
86
  "status": "pass",
87
- "value": "Recognize the current manipulation action from synchronized visual, motion, inertial, pose, and annotation context.",
88
  "raw_hits": []
89
  },
90
  {
91
- "name": "timeline_action: public_field_process_short_is_human_readable",
92
  "status": "pass",
93
- "value": "window features -> action label builder -> classifier",
94
  "raw_hits": []
95
  },
96
  {
@@ -100,9 +100,9 @@
100
  "raw_hits": []
101
  },
102
  {
103
- "name": "timeline_action: public_field_plain_goal_is_human_readable",
104
  "status": "pass",
105
- "value": "Look at one short multimodal window and name what action is happening now.",
106
  "raw_hits": []
107
  },
108
  {
@@ -184,33 +184,33 @@
184
  "observed": "timeline_subtask"
185
  },
186
  {
187
- "name": "timeline_subtask: public_field_input_short_is_human_readable",
188
  "status": "pass",
189
- "value": "20-frame multimodal window",
190
  "raw_hits": []
191
  },
192
  {
193
- "name": "timeline_subtask: public_field_research_name_is_human_readable",
194
  "status": "pass",
195
- "value": "Temporal Subtask Recognition",
196
  "raw_hits": []
197
  },
198
  {
199
- "name": "timeline_subtask: public_field_display_name_is_human_readable",
200
  "status": "pass",
201
- "value": "Procedure Step Recognition",
202
  "raw_hits": []
203
  },
204
  {
205
- "name": "timeline_subtask: public_field_card_blurb_is_human_readable",
206
  "status": "pass",
207
- "value": "Recognize the broader activity stage so fine actions become a readable procedure timeline.",
208
  "raw_hits": []
209
  },
210
  {
211
- "name": "timeline_subtask: public_field_process_short_is_human_readable",
212
  "status": "pass",
213
- "value": "window features -> subtask label builder -> classifier",
214
  "raw_hits": []
215
  },
216
  {
@@ -220,9 +220,9 @@
220
  "raw_hits": []
221
  },
222
  {
223
- "name": "timeline_subtask: public_field_plain_goal_is_human_readable",
224
  "status": "pass",
225
- "value": "Predict the higher-level task stage for the current window.",
226
  "raw_hits": []
227
  },
228
  {
@@ -304,33 +304,33 @@
304
  "observed": "transition_detection"
305
  },
306
  {
307
- "name": "transition_detection: public_field_input_short_is_human_readable",
308
  "status": "pass",
309
- "value": "current window with boundary target",
310
  "raw_hits": []
311
  },
312
  {
313
- "name": "transition_detection: public_field_research_name_is_human_readable",
314
  "status": "pass",
315
- "value": "Temporal Action Segmentation",
316
  "raw_hits": []
317
  },
318
  {
319
- "name": "transition_detection: public_field_display_name_is_human_readable",
320
  "status": "pass",
321
- "value": "Action Boundary Detection",
322
  "raw_hits": []
323
  },
324
  {
325
- "name": "transition_detection: public_field_card_blurb_is_human_readable",
326
  "status": "pass",
327
- "value": "Detect the local moment where the episode changes from one action segment to the next.",
328
  "raw_hits": []
329
  },
330
  {
331
- "name": "transition_detection: public_field_process_short_is_human_readable",
332
  "status": "pass",
333
- "value": "action changes -> boundary labels -> binary classifier",
334
  "raw_hits": []
335
  },
336
  {
@@ -340,9 +340,9 @@
340
  "raw_hits": []
341
  },
342
  {
343
- "name": "transition_detection: public_field_plain_goal_is_human_readable",
344
  "status": "pass",
345
- "value": "Detect whether the current window is near a boundary between actions.",
346
  "raw_hits": []
347
  },
348
  {
@@ -422,33 +422,33 @@
422
  "observed": "next_action"
423
  },
424
  {
425
- "name": "next_action: public_field_input_short_is_human_readable",
426
  "status": "pass",
427
- "value": "current window at time t",
428
  "raw_hits": []
429
  },
430
  {
431
- "name": "next_action: public_field_research_name_is_human_readable",
432
  "status": "pass",
433
- "value": "Short-Horizon Intention Prediction",
434
  "raw_hits": []
435
  },
436
  {
437
- "name": "next_action: public_field_display_name_is_human_readable",
438
  "status": "pass",
439
- "value": "Next-Action Prediction",
440
  "raw_hits": []
441
  },
442
  {
443
- "name": "next_action: public_field_card_blurb_is_human_readable",
444
  "status": "pass",
445
- "value": "Forecast the near-future action from the current observations only.",
446
  "raw_hits": []
447
  },
448
  {
449
- "name": "next_action: public_field_process_short_is_human_readable",
450
  "status": "pass",
451
- "value": "current features -> future label shift -> classifier",
452
  "raw_hits": []
453
  },
454
  {
@@ -458,9 +458,9 @@
458
  "raw_hits": []
459
  },
460
  {
461
- "name": "next_action: public_field_plain_goal_is_human_readable",
462
  "status": "pass",
463
- "value": "Use the current window to guess the action that will happen shortly after it.",
464
  "raw_hits": []
465
  },
466
  {
@@ -540,33 +540,33 @@
540
  "observed": "hand_trajectory_forecast"
541
  },
542
  {
543
- "name": "hand_trajectory_forecast: public_field_input_short_is_human_readable",
544
  "status": "pass",
545
- "value": "current multimodal window",
546
  "raw_hits": []
547
  },
548
  {
549
- "name": "hand_trajectory_forecast: public_field_research_name_is_human_readable",
550
  "status": "pass",
551
- "value": "3D Hand Motion Forecasting",
552
  "raw_hits": []
553
  },
554
  {
555
- "name": "hand_trajectory_forecast: public_field_display_name_is_human_readable",
556
  "status": "pass",
557
- "value": "Hand Trajectory Forecasting",
558
  "raw_hits": []
559
  },
560
  {
561
- "name": "hand_trajectory_forecast: public_field_card_blurb_is_human_readable",
562
  "status": "pass",
563
- "value": "Predict the future 3D left/right hand path from the current multimodal state.",
564
  "raw_hits": []
565
  },
566
  {
567
- "name": "hand_trajectory_forecast: public_field_process_short_is_human_readable",
568
  "status": "pass",
569
- "value": "current features -> future mocap target -> regression head",
570
  "raw_hits": []
571
  },
572
  {
@@ -576,9 +576,9 @@
576
  "raw_hits": []
577
  },
578
  {
579
- "name": "hand_trajectory_forecast: public_field_plain_goal_is_human_readable",
580
  "status": "pass",
581
- "value": "Predict where the hands will move over the next few frames.",
582
  "raw_hits": []
583
  },
584
  {
@@ -658,33 +658,33 @@
658
  "observed": "contact_prediction"
659
  },
660
  {
661
- "name": "contact_prediction: public_field_input_short_is_human_readable",
662
  "status": "pass",
663
- "value": "non-contact, non-caption features",
664
  "raw_hits": []
665
  },
666
  {
667
- "name": "contact_prediction: public_field_research_name_is_human_readable",
668
  "status": "pass",
669
- "value": "Human-Object Contact Prediction",
670
  "raw_hits": []
671
  },
672
  {
673
- "name": "contact_prediction: public_field_display_name_is_human_readable",
674
  "status": "pass",
675
- "value": "Contact State Prediction",
676
  "raw_hits": []
677
  },
678
  {
679
- "name": "contact_prediction: public_field_card_blurb_is_human_readable",
680
  "status": "pass",
681
- "value": "Predict whether body or hand contact with the scene is occurring without leaking contact labels.",
682
  "raw_hits": []
683
  },
684
  {
685
- "name": "contact_prediction: public_field_process_short_is_human_readable",
686
  "status": "pass",
687
- "value": "feature filter -> contact target -> binary classifier",
688
  "raw_hits": []
689
  },
690
  {
@@ -694,9 +694,9 @@
694
  "raw_hits": []
695
  },
696
  {
697
- "name": "contact_prediction: public_field_plain_goal_is_human_readable",
698
  "status": "pass",
699
- "value": "Predict whether the body or hand is in contact with something.",
700
  "raw_hits": []
701
  },
702
  {
@@ -774,33 +774,33 @@
774
  "observed": "object_relevance"
775
  },
776
  {
777
- "name": "object_relevance: public_field_input_short_is_human_readable",
778
  "status": "pass",
779
- "value": "non-caption multimodal features",
780
  "raw_hits": []
781
  },
782
  {
783
- "name": "object_relevance: public_field_research_name_is_human_readable",
784
  "status": "pass",
785
- "value": "Object-Centric Interaction Recognition",
786
  "raw_hits": []
787
  },
788
  {
789
- "name": "object_relevance: public_field_display_name_is_human_readable",
790
  "status": "pass",
791
- "value": "Object Relevance Prediction",
792
  "raw_hits": []
793
  },
794
  {
795
- "name": "object_relevance: public_field_card_blurb_is_human_readable",
796
  "status": "pass",
797
- "value": "Infer which objects are relevant to the current manipulation window from non-caption features.",
798
  "raw_hits": []
799
  },
800
  {
801
- "name": "object_relevance: public_field_process_short_is_human_readable",
802
  "status": "pass",
803
- "value": "object vocabulary -> multi-hot labels -> sigmoid heads",
804
  "raw_hits": []
805
  },
806
  {
@@ -810,9 +810,9 @@
810
  "raw_hits": []
811
  },
812
  {
813
- "name": "object_relevance: public_field_plain_goal_is_human_readable",
814
  "status": "pass",
815
- "value": "Predict which objects matter in the current window.",
816
  "raw_hits": []
817
  },
818
  {
@@ -892,33 +892,33 @@
892
  "observed": "caption_grounding"
893
  },
894
  {
895
- "name": "caption_grounding: public_field_input_short_is_human_readable",
896
  "status": "pass",
897
- "value": "text-like query and candidate windows",
898
  "raw_hits": []
899
  },
900
  {
901
- "name": "caption_grounding: public_field_research_name_is_human_readable",
902
  "status": "pass",
903
- "value": "Language-to-Moment Grounding",
904
  "raw_hits": []
905
  },
906
  {
907
- "name": "caption_grounding: public_field_display_name_is_human_readable",
908
  "status": "pass",
909
- "value": "Language Grounding",
910
  "raw_hits": []
911
  },
912
  {
913
- "name": "caption_grounding: public_field_card_blurb_is_human_readable",
914
  "status": "pass",
915
- "value": "Retrieve the matching time window for an annotation-derived text query.",
916
  "raw_hits": []
917
  },
918
  {
919
- "name": "caption_grounding: public_field_process_short_is_human_readable",
920
  "status": "pass",
921
- "value": "query features -> candidate index -> cosine ranker",
922
  "raw_hits": []
923
  },
924
  {
@@ -928,9 +928,9 @@
928
  "raw_hits": []
929
  },
930
  {
931
- "name": "caption_grounding: public_field_plain_goal_is_human_readable",
932
  "status": "pass",
933
- "value": "Given a text-like query from annotation, find the matching time window.",
934
  "raw_hits": []
935
  },
936
  {
@@ -1008,33 +1008,33 @@
1008
  "observed": "cross_modal_retrieval"
1009
  },
1010
  {
1011
- "name": "cross_modal_retrieval: public_field_input_short_is_human_readable",
1012
  "status": "pass",
1013
- "value": "motion/IMU/pose query; depth/video candidates",
1014
  "raw_hits": []
1015
  },
1016
  {
1017
- "name": "cross_modal_retrieval: public_field_research_name_is_human_readable",
1018
  "status": "pass",
1019
- "value": "Multimodal Representation Retrieval",
1020
  "raw_hits": []
1021
  },
1022
  {
1023
- "name": "cross_modal_retrieval: public_field_display_name_is_human_readable",
1024
  "status": "pass",
1025
- "value": "Cross-Modal Retrieval",
1026
  "raw_hits": []
1027
  },
1028
  {
1029
- "name": "cross_modal_retrieval: public_field_card_blurb_is_human_readable",
1030
  "status": "pass",
1031
- "value": "Use motion, IMU, and camera-pose signals to retrieve the matching depth/video window.",
1032
  "raw_hits": []
1033
  },
1034
  {
1035
- "name": "cross_modal_retrieval: public_field_process_short_is_human_readable",
1036
  "status": "pass",
1037
- "value": "modality split -> projection -> nearest-neighbor ranker",
1038
  "raw_hits": []
1039
  },
1040
  {
@@ -1044,9 +1044,9 @@
1044
  "raw_hits": []
1045
  },
1046
  {
1047
- "name": "cross_modal_retrieval: public_field_plain_goal_is_human_readable",
1048
  "status": "pass",
1049
- "value": "Use one group of modalities to retrieve the matching window from another group.",
1050
  "raw_hits": []
1051
  },
1052
  {
@@ -1126,33 +1126,33 @@
1126
  "observed": "modality_reconstruction"
1127
  },
1128
  {
1129
- "name": "modality_reconstruction: public_field_input_short_is_human_readable",
1130
  "status": "pass",
1131
- "value": "motion, IMU, and camera/pose features",
1132
  "raw_hits": []
1133
  },
1134
  {
1135
- "name": "modality_reconstruction: public_field_research_name_is_human_readable",
1136
  "status": "pass",
1137
- "value": "Modality Feature Reconstruction",
1138
  "raw_hits": []
1139
  },
1140
  {
1141
- "name": "modality_reconstruction: public_field_display_name_is_human_readable",
1142
  "status": "pass",
1143
- "value": "Cross-Modal Reconstruction",
1144
  "raw_hits": []
1145
  },
1146
  {
1147
- "name": "modality_reconstruction: public_field_card_blurb_is_human_readable",
1148
  "status": "pass",
1149
- "value": "Predict compressed depth/video feature vectors from motion, IMU, and camera-pose features.",
1150
  "raw_hits": []
1151
  },
1152
  {
1153
- "name": "modality_reconstruction: public_field_process_short_is_human_readable",
1154
  "status": "pass",
1155
- "value": "source-target split -> scaler -> regression head",
1156
  "raw_hits": []
1157
  },
1158
  {
@@ -1162,9 +1162,9 @@
1162
  "raw_hits": []
1163
  },
1164
  {
1165
- "name": "modality_reconstruction: public_field_plain_goal_is_human_readable",
1166
  "status": "pass",
1167
- "value": "Predict one modality feature block from other modality blocks.",
1168
  "raw_hits": []
1169
  },
1170
  {
@@ -1244,33 +1244,33 @@
1244
  "observed": "temporal_order"
1245
  },
1246
  {
1247
- "name": "temporal_order: public_field_input_short_is_human_readable",
1248
  "status": "pass",
1249
- "value": "two adjacent windows plus difference vector",
1250
  "raw_hits": []
1251
  },
1252
  {
1253
- "name": "temporal_order: public_field_research_name_is_human_readable",
1254
  "status": "pass",
1255
- "value": "Temporal Order Verification",
1256
  "raw_hits": []
1257
  },
1258
  {
1259
- "name": "temporal_order: public_field_display_name_is_human_readable",
1260
  "status": "pass",
1261
- "value": "Temporal Order Verification",
1262
  "raw_hits": []
1263
  },
1264
  {
1265
- "name": "temporal_order: public_field_card_blurb_is_human_readable",
1266
  "status": "pass",
1267
- "value": "Tell whether two neighboring windows are in chronological order or reversed.",
1268
  "raw_hits": []
1269
  },
1270
  {
1271
- "name": "temporal_order: public_field_process_short_is_human_readable",
1272
  "status": "pass",
1273
- "value": "pair builder -> feature combiner -> binary classifier",
1274
  "raw_hits": []
1275
  },
1276
  {
@@ -1280,9 +1280,9 @@
1280
  "raw_hits": []
1281
  },
1282
  {
1283
- "name": "temporal_order: public_field_plain_goal_is_human_readable",
1284
  "status": "pass",
1285
- "value": "Tell whether two nearby windows are in the correct time order.",
1286
  "raw_hits": []
1287
  },
1288
  {
@@ -1360,33 +1360,33 @@
1360
  "observed": "misalignment_detection"
1361
  },
1362
  {
1363
- "name": "misalignment_detection: public_field_input_short_is_human_readable",
1364
  "status": "pass",
1365
- "value": "motion-side and visual/depth-side feature groups",
1366
  "raw_hits": []
1367
  },
1368
  {
1369
- "name": "misalignment_detection: public_field_research_name_is_human_readable",
1370
  "status": "pass",
1371
- "value": "Cross-Modal Misalignment Detection",
1372
  "raw_hits": []
1373
  },
1374
  {
1375
- "name": "misalignment_detection: public_field_display_name_is_human_readable",
1376
  "status": "pass",
1377
- "value": "Multimodal Synchronization Detection",
1378
  "raw_hits": []
1379
  },
1380
  {
1381
- "name": "misalignment_detection: public_field_card_blurb_is_human_readable",
1382
  "status": "pass",
1383
- "value": "Detect whether motion and visual/depth streams have been artificially shifted out of sync.",
1384
  "raw_hits": []
1385
  },
1386
  {
1387
- "name": "misalignment_detection: public_field_process_short_is_human_readable",
1388
  "status": "pass",
1389
- "value": "aligned/shifted pairs -> feature combiner -> binary classifier",
1390
  "raw_hits": []
1391
  },
1392
  {
@@ -1396,9 +1396,9 @@
1396
  "raw_hits": []
1397
  },
1398
  {
1399
- "name": "misalignment_detection: public_field_plain_goal_is_human_readable",
1400
  "status": "pass",
1401
- "value": "Detect when modalities that should match are shifted out of sync.",
1402
  "raw_hits": []
1403
  },
1404
  {
 
1
  {
2
  "status": "pass",
3
+ "generated_at_utc": "2026-06-11T07:10:46+00:00",
4
  "summary": {
5
  "task_count": 12,
6
  "expected_task_count": 12,
 
64
  "observed": "timeline_action"
65
  },
66
  {
67
+ "name": "timeline_action: public_field_research_name_is_human_readable",
68
  "status": "pass",
69
+ "value": "Egocentric Action Recognition",
70
  "raw_hits": []
71
  },
72
  {
73
+ "name": "timeline_action: public_field_process_short_is_human_readable",
74
  "status": "pass",
75
+ "value": "window features -> action label builder -> classifier",
76
  "raw_hits": []
77
  },
78
  {
79
+ "name": "timeline_action: public_field_plain_goal_is_human_readable",
80
  "status": "pass",
81
+ "value": "Look at one short multimodal window and name what action is happening now.",
82
  "raw_hits": []
83
  },
84
  {
85
+ "name": "timeline_action: public_field_input_short_is_human_readable",
86
  "status": "pass",
87
+ "value": "20-frame multimodal window",
88
  "raw_hits": []
89
  },
90
  {
91
+ "name": "timeline_action: public_field_display_name_is_human_readable",
92
  "status": "pass",
93
+ "value": "Action Recognition",
94
  "raw_hits": []
95
  },
96
  {
 
100
  "raw_hits": []
101
  },
102
  {
103
+ "name": "timeline_action: public_field_card_blurb_is_human_readable",
104
  "status": "pass",
105
+ "value": "Recognize the current manipulation action from synchronized visual, motion, inertial, pose, and annotation context.",
106
  "raw_hits": []
107
  },
108
  {
 
184
  "observed": "timeline_subtask"
185
  },
186
  {
187
+ "name": "timeline_subtask: public_field_research_name_is_human_readable",
188
  "status": "pass",
189
+ "value": "Temporal Subtask Recognition",
190
  "raw_hits": []
191
  },
192
  {
193
+ "name": "timeline_subtask: public_field_process_short_is_human_readable",
194
  "status": "pass",
195
+ "value": "window features -> subtask label builder -> classifier",
196
  "raw_hits": []
197
  },
198
  {
199
+ "name": "timeline_subtask: public_field_plain_goal_is_human_readable",
200
  "status": "pass",
201
+ "value": "Predict the higher-level task stage for the current window.",
202
  "raw_hits": []
203
  },
204
  {
205
+ "name": "timeline_subtask: public_field_input_short_is_human_readable",
206
  "status": "pass",
207
+ "value": "20-frame multimodal window",
208
  "raw_hits": []
209
  },
210
  {
211
+ "name": "timeline_subtask: public_field_display_name_is_human_readable",
212
  "status": "pass",
213
+ "value": "Procedure Step Recognition",
214
  "raw_hits": []
215
  },
216
  {
 
220
  "raw_hits": []
221
  },
222
  {
223
+ "name": "timeline_subtask: public_field_card_blurb_is_human_readable",
224
  "status": "pass",
225
+ "value": "Recognize the broader activity stage so fine actions become a readable procedure timeline.",
226
  "raw_hits": []
227
  },
228
  {
 
304
  "observed": "transition_detection"
305
  },
306
  {
307
+ "name": "transition_detection: public_field_research_name_is_human_readable",
308
  "status": "pass",
309
+ "value": "Temporal Action Segmentation",
310
  "raw_hits": []
311
  },
312
  {
313
+ "name": "transition_detection: public_field_process_short_is_human_readable",
314
  "status": "pass",
315
+ "value": "action changes -> boundary labels -> binary classifier",
316
  "raw_hits": []
317
  },
318
  {
319
+ "name": "transition_detection: public_field_plain_goal_is_human_readable",
320
  "status": "pass",
321
+ "value": "Detect whether the current window is near a boundary between actions.",
322
  "raw_hits": []
323
  },
324
  {
325
+ "name": "transition_detection: public_field_input_short_is_human_readable",
326
  "status": "pass",
327
+ "value": "current window with boundary target",
328
  "raw_hits": []
329
  },
330
  {
331
+ "name": "transition_detection: public_field_display_name_is_human_readable",
332
  "status": "pass",
333
+ "value": "Action Boundary Detection",
334
  "raw_hits": []
335
  },
336
  {
 
340
  "raw_hits": []
341
  },
342
  {
343
+ "name": "transition_detection: public_field_card_blurb_is_human_readable",
344
  "status": "pass",
345
+ "value": "Detect the local moment where the episode changes from one action segment to the next.",
346
  "raw_hits": []
347
  },
348
  {
 
422
  "observed": "next_action"
423
  },
424
  {
425
+ "name": "next_action: public_field_research_name_is_human_readable",
426
  "status": "pass",
427
+ "value": "Short-Horizon Intention Prediction",
428
  "raw_hits": []
429
  },
430
  {
431
+ "name": "next_action: public_field_process_short_is_human_readable",
432
  "status": "pass",
433
+ "value": "current features -> future label shift -> classifier",
434
  "raw_hits": []
435
  },
436
  {
437
+ "name": "next_action: public_field_plain_goal_is_human_readable",
438
  "status": "pass",
439
+ "value": "Use the current window to guess the action that will happen shortly after it.",
440
  "raw_hits": []
441
  },
442
  {
443
+ "name": "next_action: public_field_input_short_is_human_readable",
444
  "status": "pass",
445
+ "value": "current window at time t",
446
  "raw_hits": []
447
  },
448
  {
449
+ "name": "next_action: public_field_display_name_is_human_readable",
450
  "status": "pass",
451
+ "value": "Next-Action Prediction",
452
  "raw_hits": []
453
  },
454
  {
 
458
  "raw_hits": []
459
  },
460
  {
461
+ "name": "next_action: public_field_card_blurb_is_human_readable",
462
  "status": "pass",
463
+ "value": "Forecast the near-future action from the current observations only.",
464
  "raw_hits": []
465
  },
466
  {
 
540
  "observed": "hand_trajectory_forecast"
541
  },
542
  {
543
+ "name": "hand_trajectory_forecast: public_field_research_name_is_human_readable",
544
  "status": "pass",
545
+ "value": "3D Hand Motion Forecasting",
546
  "raw_hits": []
547
  },
548
  {
549
+ "name": "hand_trajectory_forecast: public_field_process_short_is_human_readable",
550
  "status": "pass",
551
+ "value": "current features -> future mocap target -> regression head",
552
  "raw_hits": []
553
  },
554
  {
555
+ "name": "hand_trajectory_forecast: public_field_plain_goal_is_human_readable",
556
  "status": "pass",
557
+ "value": "Predict where the hands will move over the next few frames.",
558
  "raw_hits": []
559
  },
560
  {
561
+ "name": "hand_trajectory_forecast: public_field_input_short_is_human_readable",
562
  "status": "pass",
563
+ "value": "current multimodal window",
564
  "raw_hits": []
565
  },
566
  {
567
+ "name": "hand_trajectory_forecast: public_field_display_name_is_human_readable",
568
  "status": "pass",
569
+ "value": "Hand Trajectory Forecasting",
570
  "raw_hits": []
571
  },
572
  {
 
576
  "raw_hits": []
577
  },
578
  {
579
+ "name": "hand_trajectory_forecast: public_field_card_blurb_is_human_readable",
580
  "status": "pass",
581
+ "value": "Predict the future 3D left/right hand path from the current multimodal state.",
582
  "raw_hits": []
583
  },
584
  {
 
658
  "observed": "contact_prediction"
659
  },
660
  {
661
+ "name": "contact_prediction: public_field_research_name_is_human_readable",
662
  "status": "pass",
663
+ "value": "Human-Object Contact Prediction",
664
  "raw_hits": []
665
  },
666
  {
667
+ "name": "contact_prediction: public_field_process_short_is_human_readable",
668
  "status": "pass",
669
+ "value": "feature filter -> contact target -> binary classifier",
670
  "raw_hits": []
671
  },
672
  {
673
+ "name": "contact_prediction: public_field_plain_goal_is_human_readable",
674
  "status": "pass",
675
+ "value": "Predict whether the body or hand is in contact with something.",
676
  "raw_hits": []
677
  },
678
  {
679
+ "name": "contact_prediction: public_field_input_short_is_human_readable",
680
  "status": "pass",
681
+ "value": "non-contact, non-caption features",
682
  "raw_hits": []
683
  },
684
  {
685
+ "name": "contact_prediction: public_field_display_name_is_human_readable",
686
  "status": "pass",
687
+ "value": "Contact State Prediction",
688
  "raw_hits": []
689
  },
690
  {
 
694
  "raw_hits": []
695
  },
696
  {
697
+ "name": "contact_prediction: public_field_card_blurb_is_human_readable",
698
  "status": "pass",
699
+ "value": "Predict whether body or hand contact with the scene is occurring without leaking contact labels.",
700
  "raw_hits": []
701
  },
702
  {
 
774
  "observed": "object_relevance"
775
  },
776
  {
777
+ "name": "object_relevance: public_field_research_name_is_human_readable",
778
  "status": "pass",
779
+ "value": "Object-Centric Interaction Recognition",
780
  "raw_hits": []
781
  },
782
  {
783
+ "name": "object_relevance: public_field_process_short_is_human_readable",
784
  "status": "pass",
785
+ "value": "object vocabulary -> multi-hot labels -> sigmoid heads",
786
  "raw_hits": []
787
  },
788
  {
789
+ "name": "object_relevance: public_field_plain_goal_is_human_readable",
790
  "status": "pass",
791
+ "value": "Predict which objects matter in the current window.",
792
  "raw_hits": []
793
  },
794
  {
795
+ "name": "object_relevance: public_field_input_short_is_human_readable",
796
  "status": "pass",
797
+ "value": "non-caption multimodal features",
798
  "raw_hits": []
799
  },
800
  {
801
+ "name": "object_relevance: public_field_display_name_is_human_readable",
802
  "status": "pass",
803
+ "value": "Object Relevance Prediction",
804
  "raw_hits": []
805
  },
806
  {
 
810
  "raw_hits": []
811
  },
812
  {
813
+ "name": "object_relevance: public_field_card_blurb_is_human_readable",
814
  "status": "pass",
815
+ "value": "Infer which objects are relevant to the current manipulation window from non-caption features.",
816
  "raw_hits": []
817
  },
818
  {
 
892
  "observed": "caption_grounding"
893
  },
894
  {
895
+ "name": "caption_grounding: public_field_research_name_is_human_readable",
896
  "status": "pass",
897
+ "value": "Language-to-Moment Grounding",
898
  "raw_hits": []
899
  },
900
  {
901
+ "name": "caption_grounding: public_field_process_short_is_human_readable",
902
  "status": "pass",
903
+ "value": "query features -> candidate index -> cosine ranker",
904
  "raw_hits": []
905
  },
906
  {
907
+ "name": "caption_grounding: public_field_plain_goal_is_human_readable",
908
  "status": "pass",
909
+ "value": "Given a text-like query from annotation, find the matching time window.",
910
  "raw_hits": []
911
  },
912
  {
913
+ "name": "caption_grounding: public_field_input_short_is_human_readable",
914
  "status": "pass",
915
+ "value": "text-like query and candidate windows",
916
  "raw_hits": []
917
  },
918
  {
919
+ "name": "caption_grounding: public_field_display_name_is_human_readable",
920
  "status": "pass",
921
+ "value": "Language Grounding",
922
  "raw_hits": []
923
  },
924
  {
 
928
  "raw_hits": []
929
  },
930
  {
931
+ "name": "caption_grounding: public_field_card_blurb_is_human_readable",
932
  "status": "pass",
933
+ "value": "Retrieve the matching time window for an annotation-derived text query.",
934
  "raw_hits": []
935
  },
936
  {
 
1008
  "observed": "cross_modal_retrieval"
1009
  },
1010
  {
1011
+ "name": "cross_modal_retrieval: public_field_research_name_is_human_readable",
1012
  "status": "pass",
1013
+ "value": "Multimodal Representation Retrieval",
1014
  "raw_hits": []
1015
  },
1016
  {
1017
+ "name": "cross_modal_retrieval: public_field_process_short_is_human_readable",
1018
  "status": "pass",
1019
+ "value": "modality split -> projection -> nearest-neighbor ranker",
1020
  "raw_hits": []
1021
  },
1022
  {
1023
+ "name": "cross_modal_retrieval: public_field_plain_goal_is_human_readable",
1024
  "status": "pass",
1025
+ "value": "Use one group of modalities to retrieve the matching window from another group.",
1026
  "raw_hits": []
1027
  },
1028
  {
1029
+ "name": "cross_modal_retrieval: public_field_input_short_is_human_readable",
1030
  "status": "pass",
1031
+ "value": "motion/IMU/pose query; depth/video candidates",
1032
  "raw_hits": []
1033
  },
1034
  {
1035
+ "name": "cross_modal_retrieval: public_field_display_name_is_human_readable",
1036
  "status": "pass",
1037
+ "value": "Cross-Modal Retrieval",
1038
  "raw_hits": []
1039
  },
1040
  {
 
1044
  "raw_hits": []
1045
  },
1046
  {
1047
+ "name": "cross_modal_retrieval: public_field_card_blurb_is_human_readable",
1048
  "status": "pass",
1049
+ "value": "Use motion, IMU, and camera-pose signals to retrieve the matching depth/video window.",
1050
  "raw_hits": []
1051
  },
1052
  {
 
1126
  "observed": "modality_reconstruction"
1127
  },
1128
  {
1129
+ "name": "modality_reconstruction: public_field_research_name_is_human_readable",
1130
  "status": "pass",
1131
+ "value": "Modality Feature Reconstruction",
1132
  "raw_hits": []
1133
  },
1134
  {
1135
+ "name": "modality_reconstruction: public_field_process_short_is_human_readable",
1136
  "status": "pass",
1137
+ "value": "source-target split -> scaler -> regression head",
1138
  "raw_hits": []
1139
  },
1140
  {
1141
+ "name": "modality_reconstruction: public_field_plain_goal_is_human_readable",
1142
  "status": "pass",
1143
+ "value": "Predict one modality feature block from other modality blocks.",
1144
  "raw_hits": []
1145
  },
1146
  {
1147
+ "name": "modality_reconstruction: public_field_input_short_is_human_readable",
1148
  "status": "pass",
1149
+ "value": "motion, IMU, and camera/pose features",
1150
  "raw_hits": []
1151
  },
1152
  {
1153
+ "name": "modality_reconstruction: public_field_display_name_is_human_readable",
1154
  "status": "pass",
1155
+ "value": "Cross-Modal Reconstruction",
1156
  "raw_hits": []
1157
  },
1158
  {
 
1162
  "raw_hits": []
1163
  },
1164
  {
1165
+ "name": "modality_reconstruction: public_field_card_blurb_is_human_readable",
1166
  "status": "pass",
1167
+ "value": "Predict compressed depth/video feature vectors from motion, IMU, and camera-pose features.",
1168
  "raw_hits": []
1169
  },
1170
  {
 
1244
  "observed": "temporal_order"
1245
  },
1246
  {
1247
+ "name": "temporal_order: public_field_research_name_is_human_readable",
1248
  "status": "pass",
1249
+ "value": "Temporal Order Verification",
1250
  "raw_hits": []
1251
  },
1252
  {
1253
+ "name": "temporal_order: public_field_process_short_is_human_readable",
1254
  "status": "pass",
1255
+ "value": "pair builder -> feature combiner -> binary classifier",
1256
  "raw_hits": []
1257
  },
1258
  {
1259
+ "name": "temporal_order: public_field_plain_goal_is_human_readable",
1260
  "status": "pass",
1261
+ "value": "Tell whether two nearby windows are in the correct time order.",
1262
  "raw_hits": []
1263
  },
1264
  {
1265
+ "name": "temporal_order: public_field_input_short_is_human_readable",
1266
  "status": "pass",
1267
+ "value": "two adjacent windows plus difference vector",
1268
  "raw_hits": []
1269
  },
1270
  {
1271
+ "name": "temporal_order: public_field_display_name_is_human_readable",
1272
  "status": "pass",
1273
+ "value": "Temporal Order Verification",
1274
  "raw_hits": []
1275
  },
1276
  {
 
1280
  "raw_hits": []
1281
  },
1282
  {
1283
+ "name": "temporal_order: public_field_card_blurb_is_human_readable",
1284
  "status": "pass",
1285
+ "value": "Tell whether two neighboring windows are in chronological order or reversed.",
1286
  "raw_hits": []
1287
  },
1288
  {
 
1360
  "observed": "misalignment_detection"
1361
  },
1362
  {
1363
+ "name": "misalignment_detection: public_field_research_name_is_human_readable",
1364
  "status": "pass",
1365
+ "value": "Cross-Modal Misalignment Detection",
1366
  "raw_hits": []
1367
  },
1368
  {
1369
+ "name": "misalignment_detection: public_field_process_short_is_human_readable",
1370
  "status": "pass",
1371
+ "value": "aligned/shifted pairs -> feature combiner -> binary classifier",
1372
  "raw_hits": []
1373
  },
1374
  {
1375
+ "name": "misalignment_detection: public_field_plain_goal_is_human_readable",
1376
  "status": "pass",
1377
+ "value": "Detect when modalities that should match are shifted out of sync.",
1378
  "raw_hits": []
1379
  },
1380
  {
1381
+ "name": "misalignment_detection: public_field_input_short_is_human_readable",
1382
  "status": "pass",
1383
+ "value": "motion-side and visual/depth-side feature groups",
1384
  "raw_hits": []
1385
  },
1386
  {
1387
+ "name": "misalignment_detection: public_field_display_name_is_human_readable",
1388
  "status": "pass",
1389
+ "value": "Multimodal Synchronization Detection",
1390
  "raw_hits": []
1391
  },
1392
  {
 
1396
  "raw_hits": []
1397
  },
1398
  {
1399
+ "name": "misalignment_detection: public_field_card_blurb_is_human_readable",
1400
  "status": "pass",
1401
+ "value": "Detect whether motion and visual/depth streams have been artificially shifted out of sync.",
1402
  "raw_hits": []
1403
  },
1404
  {
data/website_integrity.json CHANGED
@@ -1,13 +1,13 @@
1
  {
2
  "status": "pass",
3
- "generated_at_utc": "2026-06-07T17:27:17+00:00",
4
  "docs_root": "docs",
5
  "site_base": "/ropedia-xperience-10m-task-suite/",
6
  "summary": {
7
  "html_pages": 4,
8
- "local_references": 137,
9
- "external_reference_count": 106,
10
- "json_files": 35,
11
  "image_assets_referenced": 22,
12
  "failure_count": 0
13
  },
@@ -75,7 +75,7 @@
75
  "status": "pass",
76
  "reason": "The project overview should appear before the deeper progress ledger.",
77
  "overview_index": 67412,
78
- "evidence_index": 90659
79
  },
80
  {
81
  "name": "project_status_links_json",
@@ -125,7 +125,7 @@
125
  "name": "roadmap_html_matches_json_phases",
126
  "status": "pass",
127
  "reason": "The roadmap section should show every stage defined in research_roadmap.json.",
128
- "phase_count": 9,
129
  "missing_phase_names": [],
130
  "roadmap_json_error": null
131
  },
@@ -133,15 +133,16 @@
133
  "name": "roadmap_status_chips_match_json",
134
  "status": "pass",
135
  "reason": "The roadmap status chips should match the phase statuses in research_roadmap.json.",
136
- "phase_count": 9,
137
  "statuses": [
138
  "implemented",
139
  "implemented_for_first_pilot",
140
  "verified_baseline",
141
  "verified_companion_result",
 
142
  "active_next_step",
143
- "next",
144
- "planned",
145
  "planned",
146
  "future"
147
  ],
@@ -153,8 +154,8 @@
153
  "status": "pass",
154
  "reason": "The evaluation protocol should appear before the deeper evidence ledger.",
155
  "overview_index": 67412,
156
- "protocol_index": 87218,
157
- "evidence_index": 90659
158
  },
159
  {
160
  "name": "evaluation_protocol_links_json",
@@ -228,7 +229,7 @@
228
  {
229
  "path": "index.html",
230
  "id_count": 77,
231
- "reference_count": 114,
232
  "image_count": 24
233
  },
234
  {
@@ -252,7 +253,7 @@
252
  },
253
  {
254
  "path": "data/artifact_index.json",
255
- "bytes": 67401,
256
  "top_level_type": "dict"
257
  },
258
  {
@@ -282,17 +283,17 @@
282
  },
283
  {
284
  "path": "data/foundation_model_plan.json",
285
- "bytes": 13193,
286
  "top_level_type": "dict"
287
  },
288
  {
289
  "path": "data/live_publication_status.json",
290
- "bytes": 74880,
291
  "top_level_type": "dict"
292
  },
293
  {
294
  "path": "data/mirror_parity.json",
295
- "bytes": 319291,
296
  "top_level_type": "dict"
297
  },
298
  {
@@ -302,12 +303,12 @@
302
  },
303
  {
304
  "path": "data/omni_finetune_verified_result.json",
305
- "bytes": 3768,
306
  "top_level_type": "dict"
307
  },
308
  {
309
  "path": "data/omni_model_comparison.json",
310
- "bytes": 50422,
311
  "top_level_type": "dict"
312
  },
313
  {
@@ -322,27 +323,32 @@
322
  },
323
  {
324
  "path": "data/project_packet.json",
325
- "bytes": 8098,
326
  "top_level_type": "dict"
327
  },
328
  {
329
  "path": "data/project_status.json",
330
- "bytes": 18062,
331
  "top_level_type": "dict"
332
  },
333
  {
334
  "path": "data/public_surface_qa.json",
335
- "bytes": 5591,
336
  "top_level_type": "dict"
337
  },
338
  {
339
  "path": "data/publication_audit.json",
340
- "bytes": 7212,
341
  "top_level_type": "dict"
342
  },
343
  {
344
  "path": "data/quality_gates.json",
345
- "bytes": 8099,
 
 
 
 
 
346
  "top_level_type": "dict"
347
  },
348
  {
@@ -367,12 +373,12 @@
367
  },
368
  {
369
  "path": "data/research_roadmap.json",
370
- "bytes": 10246,
371
  "top_level_type": "dict"
372
  },
373
  {
374
  "path": "data/research_roadmap_interactive.json",
375
- "bytes": 143673,
376
  "top_level_type": "dict"
377
  },
378
  {
@@ -382,7 +388,7 @@
382
  },
383
  {
384
  "path": "data/scope_claims_audit.json",
385
- "bytes": 21251,
386
  "top_level_type": "dict"
387
  },
388
  {
@@ -397,7 +403,12 @@
397
  },
398
  {
399
  "path": "data/summary_metrics.json",
400
- "bytes": 27490,
 
 
 
 
 
401
  "top_level_type": "dict"
402
  },
403
  {
@@ -412,7 +423,7 @@
412
  },
413
  {
414
  "path": "data/website_integrity.json",
415
- "bytes": 15375,
416
  "top_level_type": "dict"
417
  },
418
  {
 
1
  {
2
  "status": "pass",
3
+ "generated_at_utc": "2026-06-11T07:11:21+00:00",
4
  "docs_root": "docs",
5
  "site_base": "/ropedia-xperience-10m-task-suite/",
6
  "summary": {
7
  "html_pages": 4,
8
+ "local_references": 142,
9
+ "external_reference_count": 111,
10
+ "json_files": 37,
11
  "image_assets_referenced": 22,
12
  "failure_count": 0
13
  },
 
75
  "status": "pass",
76
  "reason": "The project overview should appear before the deeper progress ledger.",
77
  "overview_index": 67412,
78
+ "evidence_index": 93026
79
  },
80
  {
81
  "name": "project_status_links_json",
 
125
  "name": "roadmap_html_matches_json_phases",
126
  "status": "pass",
127
  "reason": "The roadmap section should show every stage defined in research_roadmap.json.",
128
+ "phase_count": 10,
129
  "missing_phase_names": [],
130
  "roadmap_json_error": null
131
  },
 
133
  "name": "roadmap_status_chips_match_json",
134
  "status": "pass",
135
  "reason": "The roadmap status chips should match the phase statuses in research_roadmap.json.",
136
+ "phase_count": 10,
137
  "statuses": [
138
  "implemented",
139
  "implemented_for_first_pilot",
140
  "verified_baseline",
141
  "verified_companion_result",
142
+ "current",
143
  "active_next_step",
144
+ "current",
145
+ "partially_implemented",
146
  "planned",
147
  "future"
148
  ],
 
154
  "status": "pass",
155
  "reason": "The evaluation protocol should appear before the deeper evidence ledger.",
156
  "overview_index": 67412,
157
+ "protocol_index": 89208,
158
+ "evidence_index": 93026
159
  },
160
  {
161
  "name": "evaluation_protocol_links_json",
 
229
  {
230
  "path": "index.html",
231
  "id_count": 77,
232
+ "reference_count": 119,
233
  "image_count": 24
234
  },
235
  {
 
253
  },
254
  {
255
  "path": "data/artifact_index.json",
256
+ "bytes": 83770,
257
  "top_level_type": "dict"
258
  },
259
  {
 
283
  },
284
  {
285
  "path": "data/foundation_model_plan.json",
286
+ "bytes": 13457,
287
  "top_level_type": "dict"
288
  },
289
  {
290
  "path": "data/live_publication_status.json",
291
+ "bytes": 97776,
292
  "top_level_type": "dict"
293
  },
294
  {
295
  "path": "data/mirror_parity.json",
296
+ "bytes": 521789,
297
  "top_level_type": "dict"
298
  },
299
  {
 
303
  },
304
  {
305
  "path": "data/omni_finetune_verified_result.json",
306
+ "bytes": 4431,
307
  "top_level_type": "dict"
308
  },
309
  {
310
  "path": "data/omni_model_comparison.json",
311
+ "bytes": 75707,
312
  "top_level_type": "dict"
313
  },
314
  {
 
323
  },
324
  {
325
  "path": "data/project_packet.json",
326
+ "bytes": 10496,
327
  "top_level_type": "dict"
328
  },
329
  {
330
  "path": "data/project_status.json",
331
+ "bytes": 23198,
332
  "top_level_type": "dict"
333
  },
334
  {
335
  "path": "data/public_surface_qa.json",
336
+ "bytes": 5640,
337
  "top_level_type": "dict"
338
  },
339
  {
340
  "path": "data/publication_audit.json",
341
+ "bytes": 8785,
342
  "top_level_type": "dict"
343
  },
344
  {
345
  "path": "data/quality_gates.json",
346
+ "bytes": 8097,
347
+ "top_level_type": "dict"
348
+ },
349
+ {
350
+ "path": "data/qwen3_full_parameter_gates.json",
351
+ "bytes": 10533,
352
  "top_level_type": "dict"
353
  },
354
  {
 
373
  },
374
  {
375
  "path": "data/research_roadmap.json",
376
+ "bytes": 13313,
377
  "top_level_type": "dict"
378
  },
379
  {
380
  "path": "data/research_roadmap_interactive.json",
381
+ "bytes": 145083,
382
  "top_level_type": "dict"
383
  },
384
  {
 
388
  },
389
  {
390
  "path": "data/scope_claims_audit.json",
391
+ "bytes": 21325,
392
  "top_level_type": "dict"
393
  },
394
  {
 
403
  },
404
  {
405
  "path": "data/summary_metrics.json",
406
+ "bytes": 27604,
407
+ "top_level_type": "dict"
408
+ },
409
+ {
410
+ "path": "data/task_suite_enhancement_128.json",
411
+ "bytes": 20181,
412
  "top_level_type": "dict"
413
  },
414
  {
 
423
  },
424
  {
425
  "path": "data/website_integrity.json",
426
+ "bytes": 15654,
427
  "top_level_type": "dict"
428
  },
429
  {
docs/data/mirror_parity.json CHANGED
The diff for this file is too large to render. See raw diff
 
docs/data/public_surface_qa.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "title": "Ropedia Xperience-10M Public Project Surface",
3
  "status": "pass",
4
- "generated_at_utc": "2026-06-11T03:40:05+00:00",
5
  "scope": "Repo README, GitHub Pages HTML, Hugging Face Space card, artifact dataset card, and model card.",
6
  "checks": [
7
  {
@@ -18,7 +18,7 @@
18
  "website_integrity": {
19
  "exists": true,
20
  "status": "pass",
21
- "generated_at_utc": "2026-06-11T03:23:13+00:00"
22
  },
23
  "rendered_site_check": {
24
  "exists": true,
@@ -28,27 +28,27 @@
28
  "task_surface_integrity": {
29
  "exists": true,
30
  "status": "pass",
31
- "generated_at_utc": "2026-06-11T03:36:07+00:00"
32
  },
33
  "source_alignment": {
34
  "exists": true,
35
  "status": "pass",
36
- "generated_at_utc": "2026-06-11T03:36:07+00:00"
37
  },
38
  "scale_up_status": {
39
  "exists": true,
40
  "status": "pass",
41
- "generated_at_utc": "2026-06-11T03:36:15+00:00"
42
  },
43
  "publication_package": {
44
  "exists": true,
45
  "status": "pass",
46
- "generated_at_utc": "2026-06-11T03:36:31+00:00"
47
  },
48
  "mirror_parity": {
49
  "exists": true,
50
  "status": "pass",
51
- "generated_at_utc": "2026-06-11T03:23:48+00:00"
52
  },
53
  "live_publication": {
54
  "exists": true,
 
1
  {
2
  "title": "Ropedia Xperience-10M Public Project Surface",
3
  "status": "pass",
4
+ "generated_at_utc": "2026-06-11T07:16:16+00:00",
5
  "scope": "Repo README, GitHub Pages HTML, Hugging Face Space card, artifact dataset card, and model card.",
6
  "checks": [
7
  {
 
18
  "website_integrity": {
19
  "exists": true,
20
  "status": "pass",
21
+ "generated_at_utc": "2026-06-11T07:11:21+00:00"
22
  },
23
  "rendered_site_check": {
24
  "exists": true,
 
28
  "task_surface_integrity": {
29
  "exists": true,
30
  "status": "pass",
31
+ "generated_at_utc": "2026-06-11T07:10:46+00:00"
32
  },
33
  "source_alignment": {
34
  "exists": true,
35
  "status": "pass",
36
+ "generated_at_utc": "2026-06-11T07:10:46+00:00"
37
  },
38
  "scale_up_status": {
39
  "exists": true,
40
  "status": "pass",
41
+ "generated_at_utc": "2026-06-11T07:10:47+00:00"
42
  },
43
  "publication_package": {
44
  "exists": true,
45
  "status": "pass",
46
+ "generated_at_utc": "2026-06-11T07:16:04+00:00"
47
  },
48
  "mirror_parity": {
49
  "exists": true,
50
  "status": "pass",
51
+ "generated_at_utc": "2026-06-11T07:10:20+00:00"
52
  },
53
  "live_publication": {
54
  "exists": true,
docs/data/publication_audit.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "status": "pass",
3
- "generated_at_utc": "2026-06-11T04:43:37+00:00",
4
  "checks": [
5
  {
6
  "name": "required_publication_assets_present",
@@ -209,8 +209,8 @@
209
  "hf_artifact_bundle": {
210
  "root": "hf_publish/artifacts",
211
  "exists": true,
212
- "file_count": 875,
213
- "text_file_count": 724,
214
  "largest_file": {
215
  "path": "results/episode_task_suite/modality_reconstruction/predictions.npz",
216
  "bytes": 55702978
@@ -220,8 +220,8 @@
220
  "hf_model_bundle": {
221
  "root": "hf_publish/model",
222
  "exists": true,
223
- "file_count": 1063,
224
- "text_file_count": 877,
225
  "largest_file": {
226
  "path": "pytorch_model.bin",
227
  "bytes": 93495480
 
1
  {
2
  "status": "pass",
3
+ "generated_at_utc": "2026-06-11T07:16:04+00:00",
4
  "checks": [
5
  {
6
  "name": "required_publication_assets_present",
 
209
  "hf_artifact_bundle": {
210
  "root": "hf_publish/artifacts",
211
  "exists": true,
212
+ "file_count": 877,
213
+ "text_file_count": 726,
214
  "largest_file": {
215
  "path": "results/episode_task_suite/modality_reconstruction/predictions.npz",
216
  "bytes": 55702978
 
220
  "hf_model_bundle": {
221
  "root": "hf_publish/model",
222
  "exists": true,
223
+ "file_count": 1067,
224
+ "text_file_count": 881,
225
  "largest_file": {
226
  "path": "pytorch_model.bin",
227
  "bytes": 93495480
docs/data/quality_gates.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "title": "Ropedia Xperience-10M Release Checks",
3
  "status": "pass",
4
- "generated_at_utc": "2026-06-11T03:40:04+00:00",
5
  "rule": "A release is current when the automated reports pass and the live GitHub/Hugging Face mirrors are verified after publishing.",
6
  "automated_gates": [
7
  {
 
1
  {
2
  "title": "Ropedia Xperience-10M Release Checks",
3
  "status": "pass",
4
+ "generated_at_utc": "2026-06-11T07:16:22+00:00",
5
  "rule": "A release is current when the automated reports pass and the live GitHub/Hugging Face mirrors are verified after publishing.",
6
  "automated_gates": [
7
  {
docs/data/scope_claims_audit.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "status": "pass",
3
- "generated_at_utc": "2026-06-11T03:36:15+00:00",
4
  "summary": {
5
  "qwen3_omni_verified_diagnostic_pilot": true,
6
  "dataset_manifest_num_episodes": 119,
 
1
  {
2
  "status": "pass",
3
+ "generated_at_utc": "2026-06-11T07:10:47+00:00",
4
  "summary": {
5
  "qwen3_omni_verified_diagnostic_pilot": true,
6
  "dataset_manifest_num_episodes": 119,
docs/data/source_alignment_audit.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "title": "Ropedia Xperience-10M Source Alignment Note",
3
  "status": "pass",
4
- "generated_at_utc": "2026-06-11T03:36:07+00:00",
5
  "alignment_json": "docs/data/xperience10m_dataset_card_alignment.json",
6
  "alignment_summary": {
7
  "full_dataset_repo": "ropedia-ai/xperience-10m",
 
1
  {
2
  "title": "Ropedia Xperience-10M Source Alignment Note",
3
  "status": "pass",
4
+ "generated_at_utc": "2026-06-11T07:10:46+00:00",
5
  "alignment_json": "docs/data/xperience10m_dataset_card_alignment.json",
6
  "alignment_summary": {
7
  "full_dataset_repo": "ropedia-ai/xperience-10m",
docs/data/task_surface_integrity.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "status": "pass",
3
- "generated_at_utc": "2026-06-11T03:36:07+00:00",
4
  "summary": {
5
  "task_count": 12,
6
  "expected_task_count": 12,
@@ -64,9 +64,9 @@
64
  "observed": "timeline_action"
65
  },
66
  {
67
- "name": "timeline_action: public_field_plain_goal_is_human_readable",
68
  "status": "pass",
69
- "value": "Look at one short multimodal window and name what action is happening now.",
70
  "raw_hits": []
71
  },
72
  {
@@ -76,33 +76,33 @@
76
  "raw_hits": []
77
  },
78
  {
79
- "name": "timeline_action: public_field_card_blurb_is_human_readable",
80
  "status": "pass",
81
- "value": "Recognize the current manipulation action from synchronized visual, motion, inertial, pose, and annotation context.",
82
  "raw_hits": []
83
  },
84
  {
85
- "name": "timeline_action: public_field_display_name_is_human_readable",
86
  "status": "pass",
87
- "value": "Action Recognition",
88
  "raw_hits": []
89
  },
90
  {
91
- "name": "timeline_action: public_field_research_name_is_human_readable",
92
  "status": "pass",
93
- "value": "Egocentric Action Recognition",
94
  "raw_hits": []
95
  },
96
  {
97
- "name": "timeline_action: public_field_input_short_is_human_readable",
98
  "status": "pass",
99
- "value": "20-frame multimodal window",
100
  "raw_hits": []
101
  },
102
  {
103
- "name": "timeline_action: public_field_output_short_is_human_readable",
104
  "status": "pass",
105
- "value": "current action class",
106
  "raw_hits": []
107
  },
108
  {
@@ -184,9 +184,9 @@
184
  "observed": "timeline_subtask"
185
  },
186
  {
187
- "name": "timeline_subtask: public_field_plain_goal_is_human_readable",
188
  "status": "pass",
189
- "value": "Predict the higher-level task stage for the current window.",
190
  "raw_hits": []
191
  },
192
  {
@@ -196,33 +196,33 @@
196
  "raw_hits": []
197
  },
198
  {
199
- "name": "timeline_subtask: public_field_card_blurb_is_human_readable",
200
  "status": "pass",
201
- "value": "Recognize the broader activity stage so fine actions become a readable procedure timeline.",
202
  "raw_hits": []
203
  },
204
  {
205
- "name": "timeline_subtask: public_field_display_name_is_human_readable",
206
  "status": "pass",
207
- "value": "Procedure Step Recognition",
208
  "raw_hits": []
209
  },
210
  {
211
- "name": "timeline_subtask: public_field_research_name_is_human_readable",
212
  "status": "pass",
213
- "value": "Temporal Subtask Recognition",
214
  "raw_hits": []
215
  },
216
  {
217
- "name": "timeline_subtask: public_field_input_short_is_human_readable",
218
  "status": "pass",
219
- "value": "20-frame multimodal window",
220
  "raw_hits": []
221
  },
222
  {
223
- "name": "timeline_subtask: public_field_output_short_is_human_readable",
224
  "status": "pass",
225
- "value": "current procedure step",
226
  "raw_hits": []
227
  },
228
  {
@@ -304,9 +304,9 @@
304
  "observed": "transition_detection"
305
  },
306
  {
307
- "name": "transition_detection: public_field_plain_goal_is_human_readable",
308
  "status": "pass",
309
- "value": "Detect whether the current window is near a boundary between actions.",
310
  "raw_hits": []
311
  },
312
  {
@@ -316,33 +316,33 @@
316
  "raw_hits": []
317
  },
318
  {
319
- "name": "transition_detection: public_field_card_blurb_is_human_readable",
320
  "status": "pass",
321
- "value": "Detect the local moment where the episode changes from one action segment to the next.",
322
  "raw_hits": []
323
  },
324
  {
325
- "name": "transition_detection: public_field_display_name_is_human_readable",
326
  "status": "pass",
327
- "value": "Action Boundary Detection",
328
  "raw_hits": []
329
  },
330
  {
331
- "name": "transition_detection: public_field_research_name_is_human_readable",
332
  "status": "pass",
333
- "value": "Temporal Action Segmentation",
334
  "raw_hits": []
335
  },
336
  {
337
- "name": "transition_detection: public_field_input_short_is_human_readable",
338
  "status": "pass",
339
- "value": "current window with boundary target",
340
  "raw_hits": []
341
  },
342
  {
343
- "name": "transition_detection: public_field_output_short_is_human_readable",
344
  "status": "pass",
345
- "value": "boundary or steady",
346
  "raw_hits": []
347
  },
348
  {
@@ -422,9 +422,9 @@
422
  "observed": "next_action"
423
  },
424
  {
425
- "name": "next_action: public_field_plain_goal_is_human_readable",
426
  "status": "pass",
427
- "value": "Use the current window to guess the action that will happen shortly after it.",
428
  "raw_hits": []
429
  },
430
  {
@@ -434,33 +434,33 @@
434
  "raw_hits": []
435
  },
436
  {
437
- "name": "next_action: public_field_card_blurb_is_human_readable",
438
  "status": "pass",
439
- "value": "Forecast the near-future action from the current observations only.",
440
  "raw_hits": []
441
  },
442
  {
443
- "name": "next_action: public_field_display_name_is_human_readable",
444
  "status": "pass",
445
- "value": "Next-Action Prediction",
446
  "raw_hits": []
447
  },
448
  {
449
- "name": "next_action: public_field_research_name_is_human_readable",
450
  "status": "pass",
451
- "value": "Short-Horizon Intention Prediction",
452
  "raw_hits": []
453
  },
454
  {
455
- "name": "next_action: public_field_input_short_is_human_readable",
456
  "status": "pass",
457
- "value": "current window at time t",
458
  "raw_hits": []
459
  },
460
  {
461
- "name": "next_action: public_field_output_short_is_human_readable",
462
  "status": "pass",
463
- "value": "action at t+20 frames",
464
  "raw_hits": []
465
  },
466
  {
@@ -540,9 +540,9 @@
540
  "observed": "hand_trajectory_forecast"
541
  },
542
  {
543
- "name": "hand_trajectory_forecast: public_field_plain_goal_is_human_readable",
544
  "status": "pass",
545
- "value": "Predict where the hands will move over the next few frames.",
546
  "raw_hits": []
547
  },
548
  {
@@ -552,33 +552,33 @@
552
  "raw_hits": []
553
  },
554
  {
555
- "name": "hand_trajectory_forecast: public_field_card_blurb_is_human_readable",
556
  "status": "pass",
557
- "value": "Predict the future 3D left/right hand path from the current multimodal state.",
558
  "raw_hits": []
559
  },
560
  {
561
- "name": "hand_trajectory_forecast: public_field_display_name_is_human_readable",
562
  "status": "pass",
563
- "value": "Hand Trajectory Forecasting",
564
  "raw_hits": []
565
  },
566
  {
567
- "name": "hand_trajectory_forecast: public_field_research_name_is_human_readable",
568
  "status": "pass",
569
- "value": "3D Hand Motion Forecasting",
570
  "raw_hits": []
571
  },
572
  {
573
- "name": "hand_trajectory_forecast: public_field_input_short_is_human_readable",
574
  "status": "pass",
575
- "value": "current multimodal window",
576
  "raw_hits": []
577
  },
578
  {
579
- "name": "hand_trajectory_forecast: public_field_output_short_is_human_readable",
580
  "status": "pass",
581
- "value": "future hand-joint trajectory",
582
  "raw_hits": []
583
  },
584
  {
@@ -658,9 +658,9 @@
658
  "observed": "contact_prediction"
659
  },
660
  {
661
- "name": "contact_prediction: public_field_plain_goal_is_human_readable",
662
  "status": "pass",
663
- "value": "Predict whether the body or hand is in contact with something.",
664
  "raw_hits": []
665
  },
666
  {
@@ -670,33 +670,33 @@
670
  "raw_hits": []
671
  },
672
  {
673
- "name": "contact_prediction: public_field_card_blurb_is_human_readable",
674
  "status": "pass",
675
- "value": "Predict whether body or hand contact with the scene is occurring without leaking contact labels.",
676
  "raw_hits": []
677
  },
678
  {
679
- "name": "contact_prediction: public_field_display_name_is_human_readable",
680
  "status": "pass",
681
- "value": "Contact State Prediction",
682
  "raw_hits": []
683
  },
684
  {
685
- "name": "contact_prediction: public_field_research_name_is_human_readable",
686
  "status": "pass",
687
- "value": "Human-Object Contact Prediction",
688
  "raw_hits": []
689
  },
690
  {
691
- "name": "contact_prediction: public_field_input_short_is_human_readable",
692
  "status": "pass",
693
- "value": "non-contact, non-caption features",
694
  "raw_hits": []
695
  },
696
  {
697
- "name": "contact_prediction: public_field_output_short_is_human_readable",
698
  "status": "pass",
699
- "value": "contact or no contact",
700
  "raw_hits": []
701
  },
702
  {
@@ -774,9 +774,9 @@
774
  "observed": "object_relevance"
775
  },
776
  {
777
- "name": "object_relevance: public_field_plain_goal_is_human_readable",
778
  "status": "pass",
779
- "value": "Predict which objects matter in the current window.",
780
  "raw_hits": []
781
  },
782
  {
@@ -786,33 +786,33 @@
786
  "raw_hits": []
787
  },
788
  {
789
- "name": "object_relevance: public_field_card_blurb_is_human_readable",
790
  "status": "pass",
791
- "value": "Infer which objects are relevant to the current manipulation window from non-caption features.",
792
  "raw_hits": []
793
  },
794
  {
795
- "name": "object_relevance: public_field_display_name_is_human_readable",
796
  "status": "pass",
797
- "value": "Object Relevance Prediction",
798
  "raw_hits": []
799
  },
800
  {
801
- "name": "object_relevance: public_field_research_name_is_human_readable",
802
  "status": "pass",
803
- "value": "Object-Centric Interaction Recognition",
804
  "raw_hits": []
805
  },
806
  {
807
- "name": "object_relevance: public_field_input_short_is_human_readable",
808
  "status": "pass",
809
- "value": "non-caption multimodal features",
810
  "raw_hits": []
811
  },
812
  {
813
- "name": "object_relevance: public_field_output_short_is_human_readable",
814
  "status": "pass",
815
- "value": "relevant object set",
816
  "raw_hits": []
817
  },
818
  {
@@ -892,9 +892,9 @@
892
  "observed": "caption_grounding"
893
  },
894
  {
895
- "name": "caption_grounding: public_field_plain_goal_is_human_readable",
896
  "status": "pass",
897
- "value": "Given a text-like query from annotation, find the matching time window.",
898
  "raw_hits": []
899
  },
900
  {
@@ -904,33 +904,33 @@
904
  "raw_hits": []
905
  },
906
  {
907
- "name": "caption_grounding: public_field_card_blurb_is_human_readable",
908
  "status": "pass",
909
- "value": "Retrieve the matching time window for an annotation-derived text query.",
910
  "raw_hits": []
911
  },
912
  {
913
- "name": "caption_grounding: public_field_display_name_is_human_readable",
914
  "status": "pass",
915
- "value": "Language Grounding",
916
  "raw_hits": []
917
  },
918
  {
919
- "name": "caption_grounding: public_field_research_name_is_human_readable",
920
  "status": "pass",
921
- "value": "Language-to-Moment Grounding",
922
  "raw_hits": []
923
  },
924
  {
925
- "name": "caption_grounding: public_field_input_short_is_human_readable",
926
  "status": "pass",
927
- "value": "text-like query and candidate windows",
928
  "raw_hits": []
929
  },
930
  {
931
- "name": "caption_grounding: public_field_output_short_is_human_readable",
932
  "status": "pass",
933
- "value": "ranked matching moments",
934
  "raw_hits": []
935
  },
936
  {
@@ -1008,9 +1008,9 @@
1008
  "observed": "cross_modal_retrieval"
1009
  },
1010
  {
1011
- "name": "cross_modal_retrieval: public_field_plain_goal_is_human_readable",
1012
  "status": "pass",
1013
- "value": "Use one group of modalities to retrieve the matching window from another group.",
1014
  "raw_hits": []
1015
  },
1016
  {
@@ -1020,33 +1020,33 @@
1020
  "raw_hits": []
1021
  },
1022
  {
1023
- "name": "cross_modal_retrieval: public_field_card_blurb_is_human_readable",
1024
  "status": "pass",
1025
- "value": "Use motion, IMU, and camera-pose signals to retrieve the matching depth/video window.",
1026
  "raw_hits": []
1027
  },
1028
  {
1029
- "name": "cross_modal_retrieval: public_field_display_name_is_human_readable",
1030
  "status": "pass",
1031
- "value": "Cross-Modal Retrieval",
1032
  "raw_hits": []
1033
  },
1034
  {
1035
- "name": "cross_modal_retrieval: public_field_research_name_is_human_readable",
1036
  "status": "pass",
1037
- "value": "Multimodal Representation Retrieval",
1038
  "raw_hits": []
1039
  },
1040
  {
1041
- "name": "cross_modal_retrieval: public_field_input_short_is_human_readable",
1042
  "status": "pass",
1043
- "value": "motion/IMU/pose query; depth/video candidates",
1044
  "raw_hits": []
1045
  },
1046
  {
1047
- "name": "cross_modal_retrieval: public_field_output_short_is_human_readable",
1048
  "status": "pass",
1049
- "value": "ranked visual windows",
1050
  "raw_hits": []
1051
  },
1052
  {
@@ -1126,9 +1126,9 @@
1126
  "observed": "modality_reconstruction"
1127
  },
1128
  {
1129
- "name": "modality_reconstruction: public_field_plain_goal_is_human_readable",
1130
  "status": "pass",
1131
- "value": "Predict one modality feature block from other modality blocks.",
1132
  "raw_hits": []
1133
  },
1134
  {
@@ -1138,33 +1138,33 @@
1138
  "raw_hits": []
1139
  },
1140
  {
1141
- "name": "modality_reconstruction: public_field_card_blurb_is_human_readable",
1142
  "status": "pass",
1143
- "value": "Predict compressed depth/video feature vectors from motion, IMU, and camera-pose features.",
1144
  "raw_hits": []
1145
  },
1146
  {
1147
- "name": "modality_reconstruction: public_field_display_name_is_human_readable",
1148
  "status": "pass",
1149
- "value": "Cross-Modal Reconstruction",
1150
  "raw_hits": []
1151
  },
1152
  {
1153
- "name": "modality_reconstruction: public_field_research_name_is_human_readable",
1154
  "status": "pass",
1155
- "value": "Modality Feature Reconstruction",
1156
  "raw_hits": []
1157
  },
1158
  {
1159
- "name": "modality_reconstruction: public_field_input_short_is_human_readable",
1160
  "status": "pass",
1161
- "value": "motion, IMU, and camera/pose features",
1162
  "raw_hits": []
1163
  },
1164
  {
1165
- "name": "modality_reconstruction: public_field_output_short_is_human_readable",
1166
  "status": "pass",
1167
- "value": "reconstructed depth/video vector",
1168
  "raw_hits": []
1169
  },
1170
  {
@@ -1244,9 +1244,9 @@
1244
  "observed": "temporal_order"
1245
  },
1246
  {
1247
- "name": "temporal_order: public_field_plain_goal_is_human_readable",
1248
  "status": "pass",
1249
- "value": "Tell whether two nearby windows are in the correct time order.",
1250
  "raw_hits": []
1251
  },
1252
  {
@@ -1256,33 +1256,33 @@
1256
  "raw_hits": []
1257
  },
1258
  {
1259
- "name": "temporal_order: public_field_card_blurb_is_human_readable",
1260
  "status": "pass",
1261
- "value": "Tell whether two neighboring windows are in chronological order or reversed.",
1262
  "raw_hits": []
1263
  },
1264
  {
1265
- "name": "temporal_order: public_field_display_name_is_human_readable",
1266
  "status": "pass",
1267
- "value": "Temporal Order Verification",
1268
  "raw_hits": []
1269
  },
1270
  {
1271
- "name": "temporal_order: public_field_research_name_is_human_readable",
1272
  "status": "pass",
1273
  "value": "Temporal Order Verification",
1274
  "raw_hits": []
1275
  },
1276
  {
1277
- "name": "temporal_order: public_field_input_short_is_human_readable",
1278
  "status": "pass",
1279
- "value": "two adjacent windows plus difference vector",
1280
  "raw_hits": []
1281
  },
1282
  {
1283
- "name": "temporal_order: public_field_output_short_is_human_readable",
1284
  "status": "pass",
1285
- "value": "correct or reversed",
1286
  "raw_hits": []
1287
  },
1288
  {
@@ -1360,9 +1360,9 @@
1360
  "observed": "misalignment_detection"
1361
  },
1362
  {
1363
- "name": "misalignment_detection: public_field_plain_goal_is_human_readable",
1364
  "status": "pass",
1365
- "value": "Detect when modalities that should match are shifted out of sync.",
1366
  "raw_hits": []
1367
  },
1368
  {
@@ -1372,33 +1372,33 @@
1372
  "raw_hits": []
1373
  },
1374
  {
1375
- "name": "misalignment_detection: public_field_card_blurb_is_human_readable",
1376
  "status": "pass",
1377
- "value": "Detect whether motion and visual/depth streams have been artificially shifted out of sync.",
1378
  "raw_hits": []
1379
  },
1380
  {
1381
- "name": "misalignment_detection: public_field_display_name_is_human_readable",
1382
  "status": "pass",
1383
- "value": "Multimodal Synchronization Detection",
1384
  "raw_hits": []
1385
  },
1386
  {
1387
- "name": "misalignment_detection: public_field_research_name_is_human_readable",
1388
  "status": "pass",
1389
- "value": "Cross-Modal Misalignment Detection",
1390
  "raw_hits": []
1391
  },
1392
  {
1393
- "name": "misalignment_detection: public_field_input_short_is_human_readable",
1394
  "status": "pass",
1395
- "value": "motion-side and visual/depth-side feature groups",
1396
  "raw_hits": []
1397
  },
1398
  {
1399
- "name": "misalignment_detection: public_field_output_short_is_human_readable",
1400
  "status": "pass",
1401
- "value": "aligned or shifted",
1402
  "raw_hits": []
1403
  },
1404
  {
 
1
  {
2
  "status": "pass",
3
+ "generated_at_utc": "2026-06-11T07:10:46+00:00",
4
  "summary": {
5
  "task_count": 12,
6
  "expected_task_count": 12,
 
64
  "observed": "timeline_action"
65
  },
66
  {
67
+ "name": "timeline_action: public_field_research_name_is_human_readable",
68
  "status": "pass",
69
+ "value": "Egocentric Action Recognition",
70
  "raw_hits": []
71
  },
72
  {
 
76
  "raw_hits": []
77
  },
78
  {
79
+ "name": "timeline_action: public_field_plain_goal_is_human_readable",
80
  "status": "pass",
81
+ "value": "Look at one short multimodal window and name what action is happening now.",
82
  "raw_hits": []
83
  },
84
  {
85
+ "name": "timeline_action: public_field_input_short_is_human_readable",
86
  "status": "pass",
87
+ "value": "20-frame multimodal window",
88
  "raw_hits": []
89
  },
90
  {
91
+ "name": "timeline_action: public_field_display_name_is_human_readable",
92
  "status": "pass",
93
+ "value": "Action Recognition",
94
  "raw_hits": []
95
  },
96
  {
97
+ "name": "timeline_action: public_field_output_short_is_human_readable",
98
  "status": "pass",
99
+ "value": "current action class",
100
  "raw_hits": []
101
  },
102
  {
103
+ "name": "timeline_action: public_field_card_blurb_is_human_readable",
104
  "status": "pass",
105
+ "value": "Recognize the current manipulation action from synchronized visual, motion, inertial, pose, and annotation context.",
106
  "raw_hits": []
107
  },
108
  {
 
184
  "observed": "timeline_subtask"
185
  },
186
  {
187
+ "name": "timeline_subtask: public_field_research_name_is_human_readable",
188
  "status": "pass",
189
+ "value": "Temporal Subtask Recognition",
190
  "raw_hits": []
191
  },
192
  {
 
196
  "raw_hits": []
197
  },
198
  {
199
+ "name": "timeline_subtask: public_field_plain_goal_is_human_readable",
200
  "status": "pass",
201
+ "value": "Predict the higher-level task stage for the current window.",
202
  "raw_hits": []
203
  },
204
  {
205
+ "name": "timeline_subtask: public_field_input_short_is_human_readable",
206
  "status": "pass",
207
+ "value": "20-frame multimodal window",
208
  "raw_hits": []
209
  },
210
  {
211
+ "name": "timeline_subtask: public_field_display_name_is_human_readable",
212
  "status": "pass",
213
+ "value": "Procedure Step Recognition",
214
  "raw_hits": []
215
  },
216
  {
217
+ "name": "timeline_subtask: public_field_output_short_is_human_readable",
218
  "status": "pass",
219
+ "value": "current procedure step",
220
  "raw_hits": []
221
  },
222
  {
223
+ "name": "timeline_subtask: public_field_card_blurb_is_human_readable",
224
  "status": "pass",
225
+ "value": "Recognize the broader activity stage so fine actions become a readable procedure timeline.",
226
  "raw_hits": []
227
  },
228
  {
 
304
  "observed": "transition_detection"
305
  },
306
  {
307
+ "name": "transition_detection: public_field_research_name_is_human_readable",
308
  "status": "pass",
309
+ "value": "Temporal Action Segmentation",
310
  "raw_hits": []
311
  },
312
  {
 
316
  "raw_hits": []
317
  },
318
  {
319
+ "name": "transition_detection: public_field_plain_goal_is_human_readable",
320
  "status": "pass",
321
+ "value": "Detect whether the current window is near a boundary between actions.",
322
  "raw_hits": []
323
  },
324
  {
325
+ "name": "transition_detection: public_field_input_short_is_human_readable",
326
  "status": "pass",
327
+ "value": "current window with boundary target",
328
  "raw_hits": []
329
  },
330
  {
331
+ "name": "transition_detection: public_field_display_name_is_human_readable",
332
  "status": "pass",
333
+ "value": "Action Boundary Detection",
334
  "raw_hits": []
335
  },
336
  {
337
+ "name": "transition_detection: public_field_output_short_is_human_readable",
338
  "status": "pass",
339
+ "value": "boundary or steady",
340
  "raw_hits": []
341
  },
342
  {
343
+ "name": "transition_detection: public_field_card_blurb_is_human_readable",
344
  "status": "pass",
345
+ "value": "Detect the local moment where the episode changes from one action segment to the next.",
346
  "raw_hits": []
347
  },
348
  {
 
422
  "observed": "next_action"
423
  },
424
  {
425
+ "name": "next_action: public_field_research_name_is_human_readable",
426
  "status": "pass",
427
+ "value": "Short-Horizon Intention Prediction",
428
  "raw_hits": []
429
  },
430
  {
 
434
  "raw_hits": []
435
  },
436
  {
437
+ "name": "next_action: public_field_plain_goal_is_human_readable",
438
  "status": "pass",
439
+ "value": "Use the current window to guess the action that will happen shortly after it.",
440
  "raw_hits": []
441
  },
442
  {
443
+ "name": "next_action: public_field_input_short_is_human_readable",
444
  "status": "pass",
445
+ "value": "current window at time t",
446
  "raw_hits": []
447
  },
448
  {
449
+ "name": "next_action: public_field_display_name_is_human_readable",
450
  "status": "pass",
451
+ "value": "Next-Action Prediction",
452
  "raw_hits": []
453
  },
454
  {
455
+ "name": "next_action: public_field_output_short_is_human_readable",
456
  "status": "pass",
457
+ "value": "action at t+20 frames",
458
  "raw_hits": []
459
  },
460
  {
461
+ "name": "next_action: public_field_card_blurb_is_human_readable",
462
  "status": "pass",
463
+ "value": "Forecast the near-future action from the current observations only.",
464
  "raw_hits": []
465
  },
466
  {
 
540
  "observed": "hand_trajectory_forecast"
541
  },
542
  {
543
+ "name": "hand_trajectory_forecast: public_field_research_name_is_human_readable",
544
  "status": "pass",
545
+ "value": "3D Hand Motion Forecasting",
546
  "raw_hits": []
547
  },
548
  {
 
552
  "raw_hits": []
553
  },
554
  {
555
+ "name": "hand_trajectory_forecast: public_field_plain_goal_is_human_readable",
556
  "status": "pass",
557
+ "value": "Predict where the hands will move over the next few frames.",
558
  "raw_hits": []
559
  },
560
  {
561
+ "name": "hand_trajectory_forecast: public_field_input_short_is_human_readable",
562
  "status": "pass",
563
+ "value": "current multimodal window",
564
  "raw_hits": []
565
  },
566
  {
567
+ "name": "hand_trajectory_forecast: public_field_display_name_is_human_readable",
568
  "status": "pass",
569
+ "value": "Hand Trajectory Forecasting",
570
  "raw_hits": []
571
  },
572
  {
573
+ "name": "hand_trajectory_forecast: public_field_output_short_is_human_readable",
574
  "status": "pass",
575
+ "value": "future hand-joint trajectory",
576
  "raw_hits": []
577
  },
578
  {
579
+ "name": "hand_trajectory_forecast: public_field_card_blurb_is_human_readable",
580
  "status": "pass",
581
+ "value": "Predict the future 3D left/right hand path from the current multimodal state.",
582
  "raw_hits": []
583
  },
584
  {
 
658
  "observed": "contact_prediction"
659
  },
660
  {
661
+ "name": "contact_prediction: public_field_research_name_is_human_readable",
662
  "status": "pass",
663
+ "value": "Human-Object Contact Prediction",
664
  "raw_hits": []
665
  },
666
  {
 
670
  "raw_hits": []
671
  },
672
  {
673
+ "name": "contact_prediction: public_field_plain_goal_is_human_readable",
674
  "status": "pass",
675
+ "value": "Predict whether the body or hand is in contact with something.",
676
  "raw_hits": []
677
  },
678
  {
679
+ "name": "contact_prediction: public_field_input_short_is_human_readable",
680
  "status": "pass",
681
+ "value": "non-contact, non-caption features",
682
  "raw_hits": []
683
  },
684
  {
685
+ "name": "contact_prediction: public_field_display_name_is_human_readable",
686
  "status": "pass",
687
+ "value": "Contact State Prediction",
688
  "raw_hits": []
689
  },
690
  {
691
+ "name": "contact_prediction: public_field_output_short_is_human_readable",
692
  "status": "pass",
693
+ "value": "contact or no contact",
694
  "raw_hits": []
695
  },
696
  {
697
+ "name": "contact_prediction: public_field_card_blurb_is_human_readable",
698
  "status": "pass",
699
+ "value": "Predict whether body or hand contact with the scene is occurring without leaking contact labels.",
700
  "raw_hits": []
701
  },
702
  {
 
774
  "observed": "object_relevance"
775
  },
776
  {
777
+ "name": "object_relevance: public_field_research_name_is_human_readable",
778
  "status": "pass",
779
+ "value": "Object-Centric Interaction Recognition",
780
  "raw_hits": []
781
  },
782
  {
 
786
  "raw_hits": []
787
  },
788
  {
789
+ "name": "object_relevance: public_field_plain_goal_is_human_readable",
790
  "status": "pass",
791
+ "value": "Predict which objects matter in the current window.",
792
  "raw_hits": []
793
  },
794
  {
795
+ "name": "object_relevance: public_field_input_short_is_human_readable",
796
  "status": "pass",
797
+ "value": "non-caption multimodal features",
798
  "raw_hits": []
799
  },
800
  {
801
+ "name": "object_relevance: public_field_display_name_is_human_readable",
802
  "status": "pass",
803
+ "value": "Object Relevance Prediction",
804
  "raw_hits": []
805
  },
806
  {
807
+ "name": "object_relevance: public_field_output_short_is_human_readable",
808
  "status": "pass",
809
+ "value": "relevant object set",
810
  "raw_hits": []
811
  },
812
  {
813
+ "name": "object_relevance: public_field_card_blurb_is_human_readable",
814
  "status": "pass",
815
+ "value": "Infer which objects are relevant to the current manipulation window from non-caption features.",
816
  "raw_hits": []
817
  },
818
  {
 
892
  "observed": "caption_grounding"
893
  },
894
  {
895
+ "name": "caption_grounding: public_field_research_name_is_human_readable",
896
  "status": "pass",
897
+ "value": "Language-to-Moment Grounding",
898
  "raw_hits": []
899
  },
900
  {
 
904
  "raw_hits": []
905
  },
906
  {
907
+ "name": "caption_grounding: public_field_plain_goal_is_human_readable",
908
  "status": "pass",
909
+ "value": "Given a text-like query from annotation, find the matching time window.",
910
  "raw_hits": []
911
  },
912
  {
913
+ "name": "caption_grounding: public_field_input_short_is_human_readable",
914
  "status": "pass",
915
+ "value": "text-like query and candidate windows",
916
  "raw_hits": []
917
  },
918
  {
919
+ "name": "caption_grounding: public_field_display_name_is_human_readable",
920
  "status": "pass",
921
+ "value": "Language Grounding",
922
  "raw_hits": []
923
  },
924
  {
925
+ "name": "caption_grounding: public_field_output_short_is_human_readable",
926
  "status": "pass",
927
+ "value": "ranked matching moments",
928
  "raw_hits": []
929
  },
930
  {
931
+ "name": "caption_grounding: public_field_card_blurb_is_human_readable",
932
  "status": "pass",
933
+ "value": "Retrieve the matching time window for an annotation-derived text query.",
934
  "raw_hits": []
935
  },
936
  {
 
1008
  "observed": "cross_modal_retrieval"
1009
  },
1010
  {
1011
+ "name": "cross_modal_retrieval: public_field_research_name_is_human_readable",
1012
  "status": "pass",
1013
+ "value": "Multimodal Representation Retrieval",
1014
  "raw_hits": []
1015
  },
1016
  {
 
1020
  "raw_hits": []
1021
  },
1022
  {
1023
+ "name": "cross_modal_retrieval: public_field_plain_goal_is_human_readable",
1024
  "status": "pass",
1025
+ "value": "Use one group of modalities to retrieve the matching window from another group.",
1026
  "raw_hits": []
1027
  },
1028
  {
1029
+ "name": "cross_modal_retrieval: public_field_input_short_is_human_readable",
1030
  "status": "pass",
1031
+ "value": "motion/IMU/pose query; depth/video candidates",
1032
  "raw_hits": []
1033
  },
1034
  {
1035
+ "name": "cross_modal_retrieval: public_field_display_name_is_human_readable",
1036
  "status": "pass",
1037
+ "value": "Cross-Modal Retrieval",
1038
  "raw_hits": []
1039
  },
1040
  {
1041
+ "name": "cross_modal_retrieval: public_field_output_short_is_human_readable",
1042
  "status": "pass",
1043
+ "value": "ranked visual windows",
1044
  "raw_hits": []
1045
  },
1046
  {
1047
+ "name": "cross_modal_retrieval: public_field_card_blurb_is_human_readable",
1048
  "status": "pass",
1049
+ "value": "Use motion, IMU, and camera-pose signals to retrieve the matching depth/video window.",
1050
  "raw_hits": []
1051
  },
1052
  {
 
1126
  "observed": "modality_reconstruction"
1127
  },
1128
  {
1129
+ "name": "modality_reconstruction: public_field_research_name_is_human_readable",
1130
  "status": "pass",
1131
+ "value": "Modality Feature Reconstruction",
1132
  "raw_hits": []
1133
  },
1134
  {
 
1138
  "raw_hits": []
1139
  },
1140
  {
1141
+ "name": "modality_reconstruction: public_field_plain_goal_is_human_readable",
1142
  "status": "pass",
1143
+ "value": "Predict one modality feature block from other modality blocks.",
1144
  "raw_hits": []
1145
  },
1146
  {
1147
+ "name": "modality_reconstruction: public_field_input_short_is_human_readable",
1148
  "status": "pass",
1149
+ "value": "motion, IMU, and camera/pose features",
1150
  "raw_hits": []
1151
  },
1152
  {
1153
+ "name": "modality_reconstruction: public_field_display_name_is_human_readable",
1154
  "status": "pass",
1155
+ "value": "Cross-Modal Reconstruction",
1156
  "raw_hits": []
1157
  },
1158
  {
1159
+ "name": "modality_reconstruction: public_field_output_short_is_human_readable",
1160
  "status": "pass",
1161
+ "value": "reconstructed depth/video vector",
1162
  "raw_hits": []
1163
  },
1164
  {
1165
+ "name": "modality_reconstruction: public_field_card_blurb_is_human_readable",
1166
  "status": "pass",
1167
+ "value": "Predict compressed depth/video feature vectors from motion, IMU, and camera-pose features.",
1168
  "raw_hits": []
1169
  },
1170
  {
 
1244
  "observed": "temporal_order"
1245
  },
1246
  {
1247
+ "name": "temporal_order: public_field_research_name_is_human_readable",
1248
  "status": "pass",
1249
+ "value": "Temporal Order Verification",
1250
  "raw_hits": []
1251
  },
1252
  {
 
1256
  "raw_hits": []
1257
  },
1258
  {
1259
+ "name": "temporal_order: public_field_plain_goal_is_human_readable",
1260
  "status": "pass",
1261
+ "value": "Tell whether two nearby windows are in the correct time order.",
1262
  "raw_hits": []
1263
  },
1264
  {
1265
+ "name": "temporal_order: public_field_input_short_is_human_readable",
1266
  "status": "pass",
1267
+ "value": "two adjacent windows plus difference vector",
1268
  "raw_hits": []
1269
  },
1270
  {
1271
+ "name": "temporal_order: public_field_display_name_is_human_readable",
1272
  "status": "pass",
1273
  "value": "Temporal Order Verification",
1274
  "raw_hits": []
1275
  },
1276
  {
1277
+ "name": "temporal_order: public_field_output_short_is_human_readable",
1278
  "status": "pass",
1279
+ "value": "correct or reversed",
1280
  "raw_hits": []
1281
  },
1282
  {
1283
+ "name": "temporal_order: public_field_card_blurb_is_human_readable",
1284
  "status": "pass",
1285
+ "value": "Tell whether two neighboring windows are in chronological order or reversed.",
1286
  "raw_hits": []
1287
  },
1288
  {
 
1360
  "observed": "misalignment_detection"
1361
  },
1362
  {
1363
+ "name": "misalignment_detection: public_field_research_name_is_human_readable",
1364
  "status": "pass",
1365
+ "value": "Cross-Modal Misalignment Detection",
1366
  "raw_hits": []
1367
  },
1368
  {
 
1372
  "raw_hits": []
1373
  },
1374
  {
1375
+ "name": "misalignment_detection: public_field_plain_goal_is_human_readable",
1376
  "status": "pass",
1377
+ "value": "Detect when modalities that should match are shifted out of sync.",
1378
  "raw_hits": []
1379
  },
1380
  {
1381
+ "name": "misalignment_detection: public_field_input_short_is_human_readable",
1382
  "status": "pass",
1383
+ "value": "motion-side and visual/depth-side feature groups",
1384
  "raw_hits": []
1385
  },
1386
  {
1387
+ "name": "misalignment_detection: public_field_display_name_is_human_readable",
1388
  "status": "pass",
1389
+ "value": "Multimodal Synchronization Detection",
1390
  "raw_hits": []
1391
  },
1392
  {
1393
+ "name": "misalignment_detection: public_field_output_short_is_human_readable",
1394
  "status": "pass",
1395
+ "value": "aligned or shifted",
1396
  "raw_hits": []
1397
  },
1398
  {
1399
+ "name": "misalignment_detection: public_field_card_blurb_is_human_readable",
1400
  "status": "pass",
1401
+ "value": "Detect whether motion and visual/depth streams have been artificially shifted out of sync.",
1402
  "raw_hits": []
1403
  },
1404
  {
docs/data/website_integrity.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "status": "pass",
3
- "generated_at_utc": "2026-06-11T04:43:29+00:00",
4
  "docs_root": "docs",
5
  "site_base": "/ropedia-xperience-10m-task-suite/",
6
  "summary": {
@@ -293,7 +293,7 @@
293
  },
294
  {
295
  "path": "data/mirror_parity.json",
296
- "bytes": 495491,
297
  "top_level_type": "dict"
298
  },
299
  {
@@ -338,7 +338,7 @@
338
  },
339
  {
340
  "path": "data/publication_audit.json",
341
- "bytes": 7566,
342
  "top_level_type": "dict"
343
  },
344
  {
 
1
  {
2
  "status": "pass",
3
+ "generated_at_utc": "2026-06-11T07:11:21+00:00",
4
  "docs_root": "docs",
5
  "site_base": "/ropedia-xperience-10m-task-suite/",
6
  "summary": {
 
293
  },
294
  {
295
  "path": "data/mirror_parity.json",
296
+ "bytes": 521789,
297
  "top_level_type": "dict"
298
  },
299
  {
 
338
  },
339
  {
340
  "path": "data/publication_audit.json",
341
+ "bytes": 8785,
342
  "top_level_type": "dict"
343
  },
344
  {
scripts/sync_hf_publish_mirrors.py CHANGED
@@ -22,6 +22,11 @@ STALE_MIRROR_FILES = [
22
  "artifacts/scripts/omni/collect_qwen3_v4_publication_artifacts.py",
23
  "model/scripts/omni/collect_qwen3_v4_publication_artifacts.py",
24
  ]
 
 
 
 
 
25
  ENHANCEMENT_MARKER = "docs/data/task_suite_enhancement_128.json"
26
  ENHANCEMENT_CARD_BLOCK = """
27
  ## 128-Episode Enhancement Pack
@@ -109,7 +114,25 @@ def main() -> int:
109
  src,
110
  [
111
  hf_root / "space/data" / filename,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
112
  hf_root / "artifacts/docs/data" / filename,
 
 
113
  hf_root / "model/metrics" / filename,
114
  ],
115
  dry_run=args.dry_run,
 
22
  "artifacts/scripts/omni/collect_qwen3_v4_publication_artifacts.py",
23
  "model/scripts/omni/collect_qwen3_v4_publication_artifacts.py",
24
  ]
25
+ GENERATED_REPORT_DATA_FILES = [
26
+ # The parity validator rewrites this report, so it is synced after checks
27
+ # rather than included in the self-referential hash parity file set.
28
+ "mirror_parity.json",
29
+ ]
30
  ENHANCEMENT_MARKER = "docs/data/task_suite_enhancement_128.json"
31
  ENHANCEMENT_CARD_BLOCK = """
32
  ## 128-Episode Enhancement Pack
 
114
  src,
115
  [
116
  hf_root / "space/data" / filename,
117
+ hf_root / "artifacts/data" / filename,
118
+ hf_root / "artifacts/docs/data" / filename,
119
+ hf_root / "model/data" / filename,
120
+ hf_root / "model/docs/data" / filename,
121
+ hf_root / "model/metrics" / filename,
122
+ ],
123
+ dry_run=args.dry_run,
124
+ )
125
+
126
+ for filename in GENERATED_REPORT_DATA_FILES:
127
+ src = ROOT / "docs/data" / filename
128
+ copied += copy_file(
129
+ src,
130
+ [
131
+ hf_root / "space/data" / filename,
132
+ hf_root / "artifacts/data" / filename,
133
  hf_root / "artifacts/docs/data" / filename,
134
+ hf_root / "model/data" / filename,
135
+ hf_root / "model/docs/data" / filename,
136
  hf_root / "model/metrics" / filename,
137
  ],
138
  dry_run=args.dry_run,
scripts/validate_mirror_parity.py CHANGED
@@ -318,7 +318,10 @@ def build_report(hf_root: Path) -> dict:
318
  ROOT / "docs/data" / filename,
319
  {
320
  "hf_space": hf_root / "space/data" / filename,
 
321
  "hf_artifacts": hf_root / "artifacts/docs/data" / filename,
 
 
322
  "hf_model": hf_root / "model/metrics" / filename,
323
  },
324
  hf_root,
 
318
  ROOT / "docs/data" / filename,
319
  {
320
  "hf_space": hf_root / "space/data" / filename,
321
+ "hf_artifacts_data": hf_root / "artifacts/data" / filename,
322
  "hf_artifacts": hf_root / "artifacts/docs/data" / filename,
323
+ "hf_model_data": hf_root / "model/data" / filename,
324
+ "hf_model_docs_data": hf_root / "model/docs/data" / filename,
325
  "hf_model": hf_root / "model/metrics" / filename,
326
  },
327
  hf_root,