AaronLeeky commited on
Commit
08d6c6d
·
verified ·
1 Parent(s): 9e2025e

Add files using upload-large-folder tool

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +257 -0
  2. benchmark/results/mixed_modalities_all_v1/jobs/run_omniretarget_humanref_full.sh +5 -0
  3. benchmark/results/mixed_modalities_all_v1/jobs/run_phc_humanref_full.sh +7 -0
  4. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/artifact_samples.jsonl +0 -0
  5. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/benchmark.json +1919 -0
  6. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/dataset_metrics.json +1694 -0
  7. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/embedding_metrics.json +80 -0
  8. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/metrics.json +1922 -0
  9. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/physical_metrics.json +24 -0
  10. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/real_motion_metrics.json +49 -0
  11. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/sample_metrics.jsonl +0 -0
  12. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/sample_rankings.json +1113 -0
  13. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/summary.md +10 -0
  14. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/tracker_executed.json +3 -0
  15. benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/transition_metrics.json +55 -0
  16. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/artifact_samples.jsonl +0 -0
  17. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/benchmark.json +265 -0
  18. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/dataset_metrics.json +89 -0
  19. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/embedding_metrics.json +80 -0
  20. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/metrics.json +268 -0
  21. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/physical_metrics.json +24 -0
  22. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/real_motion_metrics.json +49 -0
  23. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/sample_metrics.jsonl +0 -0
  24. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/sample_rankings.json +1113 -0
  25. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/samples.jsonl +0 -0
  26. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/summary.md +10 -0
  27. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/tracker_executed.json +3 -0
  28. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/transition_metrics.json +6 -0
  29. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/artifact_samples.jsonl +0 -0
  30. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/benchmark.json +2295 -0
  31. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/dataset_metrics.json +1046 -0
  32. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/embedding_metrics.json +80 -0
  33. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/merge_summary.json +17 -0
  34. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/metrics.json +2298 -0
  35. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/physical_metrics.json +24 -0
  36. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/real_motion_metrics.json +49 -0
  37. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/sample_metrics.jsonl +0 -0
  38. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/sample_rankings.json +1113 -0
  39. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/benchmark.json +297 -0
  40. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/dataset_metrics.json +68 -0
  41. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/embedding_metrics.json +16 -0
  42. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/metrics.json +352 -0
  43. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/physical_metrics.json +24 -0
  44. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/real_motion_metrics.json +16 -0
  45. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/reference_samples.jsonl +0 -0
  46. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/sample_metrics.jsonl +128 -0
  47. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/sample_rankings.json +3 -0
  48. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/samples.jsonl +0 -0
  49. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/summary.md +23 -0
  50. benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/transition_metrics.json +5 -0
.gitattributes CHANGED
@@ -33,3 +33,260 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00186.wav filter=lfs diff=lfs merge=lfs -text
37
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00116.wav filter=lfs diff=lfs merge=lfs -text
38
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00159.wav filter=lfs diff=lfs merge=lfs -text
39
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00416.wav filter=lfs diff=lfs merge=lfs -text
40
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00318.wav filter=lfs diff=lfs merge=lfs -text
41
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00371.wav filter=lfs diff=lfs merge=lfs -text
42
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00332.wav filter=lfs diff=lfs merge=lfs -text
43
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00127.wav filter=lfs diff=lfs merge=lfs -text
44
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00025.wav filter=lfs diff=lfs merge=lfs -text
45
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00486.wav filter=lfs diff=lfs merge=lfs -text
46
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00171.wav filter=lfs diff=lfs merge=lfs -text
47
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00252.wav filter=lfs diff=lfs merge=lfs -text
48
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00095.wav filter=lfs diff=lfs merge=lfs -text
49
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00211.wav filter=lfs diff=lfs merge=lfs -text
50
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00459.wav filter=lfs diff=lfs merge=lfs -text
51
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00078.wav filter=lfs diff=lfs merge=lfs -text
52
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00074.wav filter=lfs diff=lfs merge=lfs -text
53
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00144.wav filter=lfs diff=lfs merge=lfs -text
54
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00033.wav filter=lfs diff=lfs merge=lfs -text
55
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00242.wav filter=lfs diff=lfs merge=lfs -text
56
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00511.wav filter=lfs diff=lfs merge=lfs -text
57
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00031.wav filter=lfs diff=lfs merge=lfs -text
58
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00050.wav filter=lfs diff=lfs merge=lfs -text
59
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00297.wav filter=lfs diff=lfs merge=lfs -text
60
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00475.wav filter=lfs diff=lfs merge=lfs -text
61
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00155.wav filter=lfs diff=lfs merge=lfs -text
62
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00363.wav filter=lfs diff=lfs merge=lfs -text
63
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00004.wav filter=lfs diff=lfs merge=lfs -text
64
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00461.wav filter=lfs diff=lfs merge=lfs -text
65
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00253.wav filter=lfs diff=lfs merge=lfs -text
66
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00482.wav filter=lfs diff=lfs merge=lfs -text
67
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00301.wav filter=lfs diff=lfs merge=lfs -text
68
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00080.wav filter=lfs diff=lfs merge=lfs -text
69
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00406.wav filter=lfs diff=lfs merge=lfs -text
70
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00508.wav filter=lfs diff=lfs merge=lfs -text
71
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00378.wav filter=lfs diff=lfs merge=lfs -text
72
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00038.wav filter=lfs diff=lfs merge=lfs -text
73
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00015.wav filter=lfs diff=lfs merge=lfs -text
74
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00331.wav filter=lfs diff=lfs merge=lfs -text
75
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00489.wav filter=lfs diff=lfs merge=lfs -text
76
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00161.wav filter=lfs diff=lfs merge=lfs -text
77
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00056.wav filter=lfs diff=lfs merge=lfs -text
78
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00395.wav filter=lfs diff=lfs merge=lfs -text
79
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00304.wav filter=lfs diff=lfs merge=lfs -text
80
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00091.wav filter=lfs diff=lfs merge=lfs -text
81
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00140.wav filter=lfs diff=lfs merge=lfs -text
82
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00409.wav filter=lfs diff=lfs merge=lfs -text
83
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00064.wav filter=lfs diff=lfs merge=lfs -text
84
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00325.wav filter=lfs diff=lfs merge=lfs -text
85
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00032.wav filter=lfs diff=lfs merge=lfs -text
86
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00005.wav filter=lfs diff=lfs merge=lfs -text
87
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00279.wav filter=lfs diff=lfs merge=lfs -text
88
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00328.wav filter=lfs diff=lfs merge=lfs -text
89
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00137.wav filter=lfs diff=lfs merge=lfs -text
90
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00439.wav filter=lfs diff=lfs merge=lfs -text
91
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00234.wav filter=lfs diff=lfs merge=lfs -text
92
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00067.wav filter=lfs diff=lfs merge=lfs -text
93
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00310.wav filter=lfs diff=lfs merge=lfs -text
94
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00158.wav filter=lfs diff=lfs merge=lfs -text
95
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00000.wav filter=lfs diff=lfs merge=lfs -text
96
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00315.wav filter=lfs diff=lfs merge=lfs -text
97
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00283.wav filter=lfs diff=lfs merge=lfs -text
98
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00356.wav filter=lfs diff=lfs merge=lfs -text
99
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00455.wav filter=lfs diff=lfs merge=lfs -text
100
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00029.wav filter=lfs diff=lfs merge=lfs -text
101
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00199.wav filter=lfs diff=lfs merge=lfs -text
102
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00105.wav filter=lfs diff=lfs merge=lfs -text
103
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00073.wav filter=lfs diff=lfs merge=lfs -text
104
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00390.wav filter=lfs diff=lfs merge=lfs -text
105
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00275.wav filter=lfs diff=lfs merge=lfs -text
106
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00014.wav filter=lfs diff=lfs merge=lfs -text
107
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00373.wav filter=lfs diff=lfs merge=lfs -text
108
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00311.wav filter=lfs diff=lfs merge=lfs -text
109
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00422.wav filter=lfs diff=lfs merge=lfs -text
110
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00388.wav filter=lfs diff=lfs merge=lfs -text
111
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00141.wav filter=lfs diff=lfs merge=lfs -text
112
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00230.wav filter=lfs diff=lfs merge=lfs -text
113
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00233.wav filter=lfs diff=lfs merge=lfs -text
114
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00495.wav filter=lfs diff=lfs merge=lfs -text
115
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00119.wav filter=lfs diff=lfs merge=lfs -text
116
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00487.wav filter=lfs diff=lfs merge=lfs -text
117
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00203.wav filter=lfs diff=lfs merge=lfs -text
118
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00184.wav filter=lfs diff=lfs merge=lfs -text
119
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00419.wav filter=lfs diff=lfs merge=lfs -text
120
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00052.wav filter=lfs diff=lfs merge=lfs -text
121
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00042.wav filter=lfs diff=lfs merge=lfs -text
122
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00147.wav filter=lfs diff=lfs merge=lfs -text
123
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00441.wav filter=lfs diff=lfs merge=lfs -text
124
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00289.wav filter=lfs diff=lfs merge=lfs -text
125
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00448.wav filter=lfs diff=lfs merge=lfs -text
126
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00010.wav filter=lfs diff=lfs merge=lfs -text
127
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00502.wav filter=lfs diff=lfs merge=lfs -text
128
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00261.wav filter=lfs diff=lfs merge=lfs -text
129
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00366.wav filter=lfs diff=lfs merge=lfs -text
130
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00268.wav filter=lfs diff=lfs merge=lfs -text
131
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00053.wav filter=lfs diff=lfs merge=lfs -text
132
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00212.wav filter=lfs diff=lfs merge=lfs -text
133
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00474.wav filter=lfs diff=lfs merge=lfs -text
134
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00007.wav filter=lfs diff=lfs merge=lfs -text
135
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00083.wav filter=lfs diff=lfs merge=lfs -text
136
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00339.wav filter=lfs diff=lfs merge=lfs -text
137
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00498.wav filter=lfs diff=lfs merge=lfs -text
138
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00396.wav filter=lfs diff=lfs merge=lfs -text
139
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00131.wav filter=lfs diff=lfs merge=lfs -text
140
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00481.wav filter=lfs diff=lfs merge=lfs -text
141
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00418.wav filter=lfs diff=lfs merge=lfs -text
142
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00258.wav filter=lfs diff=lfs merge=lfs -text
143
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00121.wav filter=lfs diff=lfs merge=lfs -text
144
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00164.wav filter=lfs diff=lfs merge=lfs -text
145
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00226.wav filter=lfs diff=lfs merge=lfs -text
146
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00488.wav filter=lfs diff=lfs merge=lfs -text
147
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00104.wav filter=lfs diff=lfs merge=lfs -text
148
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00209.wav filter=lfs diff=lfs merge=lfs -text
149
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00432.wav filter=lfs diff=lfs merge=lfs -text
150
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00100.wav filter=lfs diff=lfs merge=lfs -text
151
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00334.wav filter=lfs diff=lfs merge=lfs -text
152
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00185.wav filter=lfs diff=lfs merge=lfs -text
153
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00400.wav filter=lfs diff=lfs merge=lfs -text
154
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00177.wav filter=lfs diff=lfs merge=lfs -text
155
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00330.wav filter=lfs diff=lfs merge=lfs -text
156
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00303.wav filter=lfs diff=lfs merge=lfs -text
157
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00284.wav filter=lfs diff=lfs merge=lfs -text
158
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00188.wav filter=lfs diff=lfs merge=lfs -text
159
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00447.wav filter=lfs diff=lfs merge=lfs -text
160
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00115.wav filter=lfs diff=lfs merge=lfs -text
161
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00118.wav filter=lfs diff=lfs merge=lfs -text
162
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00024.wav filter=lfs diff=lfs merge=lfs -text
163
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00324.wav filter=lfs diff=lfs merge=lfs -text
164
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00429.wav filter=lfs diff=lfs merge=lfs -text
165
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00320.wav filter=lfs diff=lfs merge=lfs -text
166
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00132.wav filter=lfs diff=lfs merge=lfs -text
167
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00397.wav filter=lfs diff=lfs merge=lfs -text
168
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00299.wav filter=lfs diff=lfs merge=lfs -text
169
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00365.wav filter=lfs diff=lfs merge=lfs -text
170
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00473.wav filter=lfs diff=lfs merge=lfs -text
171
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00401.wav filter=lfs diff=lfs merge=lfs -text
172
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00046.wav filter=lfs diff=lfs merge=lfs -text
173
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00241.wav filter=lfs diff=lfs merge=lfs -text
174
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00346.wav filter=lfs diff=lfs merge=lfs -text
175
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00076.wav filter=lfs diff=lfs merge=lfs -text
176
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00160.wav filter=lfs diff=lfs merge=lfs -text
177
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00097.wav filter=lfs diff=lfs merge=lfs -text
178
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00312.wav filter=lfs diff=lfs merge=lfs -text
179
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00163.wav filter=lfs diff=lfs merge=lfs -text
180
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00376.wav filter=lfs diff=lfs merge=lfs -text
181
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00036.wav filter=lfs diff=lfs merge=lfs -text
182
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00101.wav filter=lfs diff=lfs merge=lfs -text
183
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00470.wav filter=lfs diff=lfs merge=lfs -text
184
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00206.wav filter=lfs diff=lfs merge=lfs -text
185
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00178.wav filter=lfs diff=lfs merge=lfs -text
186
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00302.wav filter=lfs diff=lfs merge=lfs -text
187
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00174.wav filter=lfs diff=lfs merge=lfs -text
188
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00191.wav filter=lfs diff=lfs merge=lfs -text
189
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00157.wav filter=lfs diff=lfs merge=lfs -text
190
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00309.wav filter=lfs diff=lfs merge=lfs -text
191
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00020.wav filter=lfs diff=lfs merge=lfs -text
192
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00277.wav filter=lfs diff=lfs merge=lfs -text
193
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00379.wav filter=lfs diff=lfs merge=lfs -text
194
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00089.wav filter=lfs diff=lfs merge=lfs -text
195
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00317.wav filter=lfs diff=lfs merge=lfs -text
196
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00410.wav filter=lfs diff=lfs merge=lfs -text
197
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00298.wav filter=lfs diff=lfs merge=lfs -text
198
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00313.wav filter=lfs diff=lfs merge=lfs -text
199
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00294.wav filter=lfs diff=lfs merge=lfs -text
200
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00232.wav filter=lfs diff=lfs merge=lfs -text
201
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00345.wav filter=lfs diff=lfs merge=lfs -text
202
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00372.wav filter=lfs diff=lfs merge=lfs -text
203
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00496.wav filter=lfs diff=lfs merge=lfs -text
204
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00466.wav filter=lfs diff=lfs merge=lfs -text
205
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00392.wav filter=lfs diff=lfs merge=lfs -text
206
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00407.wav filter=lfs diff=lfs merge=lfs -text
207
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00505.wav filter=lfs diff=lfs merge=lfs -text
208
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00375.wav filter=lfs diff=lfs merge=lfs -text
209
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00061.wav filter=lfs diff=lfs merge=lfs -text
210
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00221.wav filter=lfs diff=lfs merge=lfs -text
211
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00166.wav filter=lfs diff=lfs merge=lfs -text
212
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00197.wav filter=lfs diff=lfs merge=lfs -text
213
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00156.wav filter=lfs diff=lfs merge=lfs -text
214
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00456.wav filter=lfs diff=lfs merge=lfs -text
215
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00139.wav filter=lfs diff=lfs merge=lfs -text
216
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00002.wav filter=lfs diff=lfs merge=lfs -text
217
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00264.wav filter=lfs diff=lfs merge=lfs -text
218
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00369.wav filter=lfs diff=lfs merge=lfs -text
219
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00382.wav filter=lfs diff=lfs merge=lfs -text
220
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00260.wav filter=lfs diff=lfs merge=lfs -text
221
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00190.wav filter=lfs diff=lfs merge=lfs -text
222
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00072.wav filter=lfs diff=lfs merge=lfs -text
223
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00235.wav filter=lfs diff=lfs merge=lfs -text
224
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00075.wav filter=lfs diff=lfs merge=lfs -text
225
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00218.wav filter=lfs diff=lfs merge=lfs -text
226
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00386.wav filter=lfs diff=lfs merge=lfs -text
227
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00026.wav filter=lfs diff=lfs merge=lfs -text
228
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00288.wav filter=lfs diff=lfs merge=lfs -text
229
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00354.wav filter=lfs diff=lfs merge=lfs -text
230
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00009.wav filter=lfs diff=lfs merge=lfs -text
231
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00359.wav filter=lfs diff=lfs merge=lfs -text
232
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00222.wav filter=lfs diff=lfs merge=lfs -text
233
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00424.wav filter=lfs diff=lfs merge=lfs -text
234
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00257.wav filter=lfs diff=lfs merge=lfs -text
235
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00030.wav filter=lfs diff=lfs merge=lfs -text
236
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00135.wav filter=lfs diff=lfs merge=lfs -text
237
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00287.wav filter=lfs diff=lfs merge=lfs -text
238
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00003.wav filter=lfs diff=lfs merge=lfs -text
239
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00133.wav filter=lfs diff=lfs merge=lfs -text
240
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00237.wav filter=lfs diff=lfs merge=lfs -text
241
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00179.wav filter=lfs diff=lfs merge=lfs -text
242
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00219.wav filter=lfs diff=lfs merge=lfs -text
243
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00446.wav filter=lfs diff=lfs merge=lfs -text
244
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00117.wav filter=lfs diff=lfs merge=lfs -text
245
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00476.wav filter=lfs diff=lfs merge=lfs -text
246
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00054.wav filter=lfs diff=lfs merge=lfs -text
247
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00355.wav filter=lfs diff=lfs merge=lfs -text
248
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00248.wav filter=lfs diff=lfs merge=lfs -text
249
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00085.wav filter=lfs diff=lfs merge=lfs -text
250
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00201.wav filter=lfs diff=lfs merge=lfs -text
251
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00136.wav filter=lfs diff=lfs merge=lfs -text
252
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00449.wav filter=lfs diff=lfs merge=lfs -text
253
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00266.wav filter=lfs diff=lfs merge=lfs -text
254
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00350.wav filter=lfs diff=lfs merge=lfs -text
255
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00336.wav filter=lfs diff=lfs merge=lfs -text
256
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00291.wav filter=lfs diff=lfs merge=lfs -text
257
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00068.wav filter=lfs diff=lfs merge=lfs -text
258
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00465.wav filter=lfs diff=lfs merge=lfs -text
259
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00012.wav filter=lfs diff=lfs merge=lfs -text
260
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00274.wav filter=lfs diff=lfs merge=lfs -text
261
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00152.wav filter=lfs diff=lfs merge=lfs -text
262
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00243.wav filter=lfs diff=lfs merge=lfs -text
263
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00347.wav filter=lfs diff=lfs merge=lfs -text
264
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00196.wav filter=lfs diff=lfs merge=lfs -text
265
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00417.wav filter=lfs diff=lfs merge=lfs -text
266
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00398.wav filter=lfs diff=lfs merge=lfs -text
267
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00109.wav filter=lfs diff=lfs merge=lfs -text
268
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00281.wav filter=lfs diff=lfs merge=lfs -text
269
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00134.wav filter=lfs diff=lfs merge=lfs -text
270
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00423.wav filter=lfs diff=lfs merge=lfs -text
271
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00267.wav filter=lfs diff=lfs merge=lfs -text
272
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00200.wav filter=lfs diff=lfs merge=lfs -text
273
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00434.wav filter=lfs diff=lfs merge=lfs -text
274
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00130.wav filter=lfs diff=lfs merge=lfs -text
275
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00022.wav filter=lfs diff=lfs merge=lfs -text
276
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00113.wav filter=lfs diff=lfs merge=lfs -text
277
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00103.wav filter=lfs diff=lfs merge=lfs -text
278
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00084.wav filter=lfs diff=lfs merge=lfs -text
279
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00348.wav filter=lfs diff=lfs merge=lfs -text
280
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00192.wav filter=lfs diff=lfs merge=lfs -text
281
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00469.wav filter=lfs diff=lfs merge=lfs -text
282
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00051.wav filter=lfs diff=lfs merge=lfs -text
283
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00175.wav filter=lfs diff=lfs merge=lfs -text
284
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00402.wav filter=lfs diff=lfs merge=lfs -text
285
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00099.wav filter=lfs diff=lfs merge=lfs -text
286
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00037.wav filter=lfs diff=lfs merge=lfs -text
287
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00322.wav filter=lfs diff=lfs merge=lfs -text
288
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00329.wav filter=lfs diff=lfs merge=lfs -text
289
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00165.wav filter=lfs diff=lfs merge=lfs -text
290
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00041.wav filter=lfs diff=lfs merge=lfs -text
291
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00413.wav filter=lfs diff=lfs merge=lfs -text
292
+ benchmark/samples/mixed_modalities_all_v1/.audio_120f_export/audio/sample_00208.wav filter=lfs diff=lfs merge=lfs -text
benchmark/results/mixed_modalities_all_v1/jobs/run_omniretarget_humanref_full.sh ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ #!/usr/bin/env bash
2
+ set -euo pipefail
3
+ cd ${OMG_REPO_ROOT}
4
+ HOLOSOMA_PYTHON=${OMG_REPO_ROOT}/.venv/bin/python PYTHON=.venv/bin/python bash baselines/omniretarget/run_humanref_benchmark.sh
5
+ scripts/benchmarks/update_benchmark_md.sh
benchmark/results/mixed_modalities_all_v1/jobs/run_phc_humanref_full.sh ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ #!/usr/bin/env bash
2
+ set -euo pipefail
3
+ cd ${OMG_REPO_ROOT}
4
+ export PYTHONPATH=src
5
+ .venv/bin/python scripts/benchmarks/materialize_humanref_baseline_inputs.py --stages keypoints phc_amass --num-frames 60 --device cuda
6
+ PHC_ROOT=${OMG_EXTERNAL_ROOT}/PHC PHC_PYTHON=${OMG_REPO_ROOT}/.venv/bin/python PYTHON=.venv/bin/python bash baselines/phc/run_humanref_benchmark.sh
7
+ scripts/benchmarks/update_benchmark_md.sh
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/artifact_samples.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/benchmark.json ADDED
@@ -0,0 +1,1919 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark": "artifact_text",
3
+ "dataset_metrics": {
4
+ "100style_original_filtered_test": {
5
+ "embedding": {
6
+ "diversity_generated": 1.0704150905599346,
7
+ "motion_fid": 1.2838937050444919,
8
+ "motion_kid": {
9
+ "mean": 0.0022249471798367892,
10
+ "std": 0.0
11
+ },
12
+ "num_samples": 43
13
+ },
14
+ "num_samples": 43,
15
+ "physical": {
16
+ "body_jerk_mean": {
17
+ "max": 154.36744689941406,
18
+ "mean": 73.70867042208828,
19
+ "min": 2.9786179065704346,
20
+ "std": 37.651289022141334
21
+ },
22
+ "contact_sliding_speed": {
23
+ "max": 1.0541962385177612,
24
+ "mean": 0.3344618064286404,
25
+ "min": 0.007166638039052486,
26
+ "std": 0.2880695016780662
27
+ },
28
+ "foot_ground_error": {
29
+ "max": 0.015996823087334633,
30
+ "mean": 0.005727387244574899,
31
+ "min": 0.0019035614095628262,
32
+ "std": 0.002544252817203053
33
+ },
34
+ "num_samples": 43
35
+ },
36
+ "text": {
37
+ "generated": {
38
+ "matching_distance": {
39
+ "max": 1.4084994792938232,
40
+ "mean": 1.3151017427444458,
41
+ "min": 1.2379783391952515,
42
+ "std": 0.036907895016914875
43
+ },
44
+ "matching_score": 1.3151017427444458,
45
+ "mean_text_rank": 10.90625,
46
+ "median_text_rank": 10.0,
47
+ "r_precision": [
48
+ 0.09375,
49
+ 0.125,
50
+ 0.1875
51
+ ],
52
+ "retrieval_batch_size": 32,
53
+ "retrieval_num_batches": 1,
54
+ "retrieval_num_samples": 32,
55
+ "retrieval_order_policy": "manifest_order",
56
+ "retrieval_unique_datasets_per_batch": {
57
+ "max": 1.0,
58
+ "mean": 1.0,
59
+ "min": 1.0,
60
+ "std": 0.0
61
+ }
62
+ },
63
+ "reference": {
64
+ "matching_distance": {
65
+ "max": 1.3596904277801514,
66
+ "mean": 1.26676583567331,
67
+ "min": 1.1853543519973755,
68
+ "std": 0.03781489479220778
69
+ },
70
+ "matching_score": 1.266765832901001,
71
+ "mean_text_rank": 5.71875,
72
+ "median_text_rank": 3.5,
73
+ "r_precision": [
74
+ 0.34375,
75
+ 0.46875,
76
+ 0.5
77
+ ],
78
+ "retrieval_batch_size": 32,
79
+ "retrieval_num_batches": 1,
80
+ "retrieval_num_samples": 32,
81
+ "retrieval_order_policy": "manifest_order",
82
+ "retrieval_unique_datasets_per_batch": {
83
+ "max": 1.0,
84
+ "mean": 1.0,
85
+ "min": 1.0,
86
+ "std": 0.0
87
+ }
88
+ }
89
+ },
90
+ "transition": {
91
+ "auj": {
92
+ "max": 3.0685779272303186,
93
+ "mean": 1.8053734862221296,
94
+ "min": 1.6700049626518585,
95
+ "std": 0.20044042082439859
96
+ },
97
+ "body_auj": {
98
+ "max": 3.0685779272303186,
99
+ "mean": 1.8053734862221296,
100
+ "min": 1.6700049626518585,
101
+ "std": 0.20044042082439859
102
+ },
103
+ "body_pj": {
104
+ "max": 0.6321938335895538,
105
+ "mean": 0.06189292100756321,
106
+ "min": 0.0008932948112487793,
107
+ "std": 0.09465196402988692
108
+ },
109
+ "boundary_frames": [
110
+ 60
111
+ ],
112
+ "chunk_length": 60,
113
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
114
+ "num_frames": 120,
115
+ "num_samples": 43,
116
+ "pj": {
117
+ "max": 0.6321938335895538,
118
+ "mean": 0.06189292100756321,
119
+ "min": 0.0008932948112487793,
120
+ "std": 0.09465196402988692
121
+ },
122
+ "qpos_auj": {
123
+ "max": 24.71011780718753,
124
+ "mean": 6.1841539223100535,
125
+ "min": 4.2977976540647,
126
+ "std": 3.838152676323723
127
+ },
128
+ "qpos_pj": {
129
+ "max": 9.507055640220642,
130
+ "mean": 0.9185523066323164,
131
+ "min": 0.003028661012649536,
132
+ "std": 1.6940965662424199
133
+ },
134
+ "reference_body_jerk_level": 0.032168975994990016,
135
+ "reference_qpos_jerk_level": 0.07551584170589375,
136
+ "transition_window_length": 60,
137
+ "transition_windows": [
138
+ [
139
+ 30,
140
+ 90
141
+ ]
142
+ ]
143
+ }
144
+ },
145
+ "amass_original_filtered_test": {
146
+ "embedding": {
147
+ "diversity_generated": 1.111925104779673,
148
+ "motion_fid": 1.328561012147938,
149
+ "motion_kid": {
150
+ "mean": 0.0022260718776268007,
151
+ "std": 0.0
152
+ },
153
+ "num_samples": 43
154
+ },
155
+ "num_samples": 43,
156
+ "physical": {
157
+ "body_jerk_mean": {
158
+ "max": 269.9172668457031,
159
+ "mean": 95.9706466918768,
160
+ "min": 2.5841948986053467,
161
+ "std": 70.75177739661642
162
+ },
163
+ "contact_sliding_speed": {
164
+ "max": 1.5415642261505127,
165
+ "mean": 0.39597279582779077,
166
+ "min": 0.005511525087058544,
167
+ "std": 0.40617249845569314
168
+ },
169
+ "foot_ground_error": {
170
+ "max": 0.04935852810740471,
171
+ "mean": 0.00705925366559694,
172
+ "min": 0.0024636401794850826,
173
+ "std": 0.007562361206799781
174
+ },
175
+ "num_samples": 43
176
+ },
177
+ "text": {
178
+ "generated": {
179
+ "matching_distance": {
180
+ "max": 1.418434739112854,
181
+ "mean": 1.3283776654753574,
182
+ "min": 1.2401793003082275,
183
+ "std": 0.04136207243662587
184
+ },
185
+ "matching_score": 1.328377604484558,
186
+ "mean_text_rank": 6.96875,
187
+ "median_text_rank": 3.0,
188
+ "r_precision": [
189
+ 0.375,
190
+ 0.5,
191
+ 0.5
192
+ ],
193
+ "retrieval_batch_size": 32,
194
+ "retrieval_num_batches": 1,
195
+ "retrieval_num_samples": 32,
196
+ "retrieval_order_policy": "manifest_order",
197
+ "retrieval_unique_datasets_per_batch": {
198
+ "max": 1.0,
199
+ "mean": 1.0,
200
+ "min": 1.0,
201
+ "std": 0.0
202
+ }
203
+ },
204
+ "reference": {
205
+ "matching_distance": {
206
+ "max": 1.4135501384735107,
207
+ "mean": 1.272889874702276,
208
+ "min": 1.172876000404358,
209
+ "std": 0.04160945025899305
210
+ },
211
+ "matching_score": 1.2728898525238037,
212
+ "mean_text_rank": 4.40625,
213
+ "median_text_rank": 2.0,
214
+ "r_precision": [
215
+ 0.3125,
216
+ 0.5625,
217
+ 0.625
218
+ ],
219
+ "retrieval_batch_size": 32,
220
+ "retrieval_num_batches": 1,
221
+ "retrieval_num_samples": 32,
222
+ "retrieval_order_policy": "manifest_order",
223
+ "retrieval_unique_datasets_per_batch": {
224
+ "max": 1.0,
225
+ "mean": 1.0,
226
+ "min": 1.0,
227
+ "std": 0.0
228
+ }
229
+ }
230
+ },
231
+ "transition": {
232
+ "auj": {
233
+ "max": 6.514222457101631,
234
+ "mean": 1.8822794797577118,
235
+ "min": 1.5973546691528981,
236
+ "std": 0.7210500357139227
237
+ },
238
+ "body_auj": {
239
+ "max": 6.514222457101631,
240
+ "mean": 1.8822794797577118,
241
+ "min": 1.5973546691528981,
242
+ "std": 0.7210500357139227
243
+ },
244
+ "body_pj": {
245
+ "max": 1.9185047969222069,
246
+ "mean": 0.09840007811828061,
247
+ "min": 0.00058785080909729,
248
+ "std": 0.28364867706203306
249
+ },
250
+ "boundary_frames": [
251
+ 60
252
+ ],
253
+ "chunk_length": 60,
254
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
255
+ "num_frames": 120,
256
+ "num_samples": 43,
257
+ "pj": {
258
+ "max": 1.9185047969222069,
259
+ "mean": 0.09840007811828061,
260
+ "min": 0.00058785080909729,
261
+ "std": 0.28364867706203306
262
+ },
263
+ "qpos_auj": {
264
+ "max": 72.83266460476783,
265
+ "mean": 8.4905600420132,
266
+ "min": 4.2940439462623115,
267
+ "std": 11.744696961776553
268
+ },
269
+ "qpos_pj": {
270
+ "max": 12.584792256355286,
271
+ "mean": 1.159780290310791,
272
+ "min": 0.002947501838207245,
273
+ "std": 2.334767403530377
274
+ },
275
+ "reference_body_jerk_level": 0.032168975994990016,
276
+ "reference_qpos_jerk_level": 0.07551584170589375,
277
+ "transition_window_length": 60,
278
+ "transition_windows": [
279
+ [
280
+ 30,
281
+ 90
282
+ ]
283
+ ]
284
+ }
285
+ },
286
+ "bones_seed_original_filtered_test": {
287
+ "embedding": {
288
+ "diversity_generated": 1.1937374883617156,
289
+ "motion_fid": 0.4045952532270245,
290
+ "motion_kid": {
291
+ "mean": -0.00010609229304492374,
292
+ "std": 0.0
293
+ },
294
+ "num_samples": 43
295
+ },
296
+ "num_samples": 43,
297
+ "physical": {
298
+ "body_jerk_mean": {
299
+ "max": 261.4084777832031,
300
+ "mean": 69.5281113635662,
301
+ "min": 3.05818772315979,
302
+ "std": 68.85240862628845
303
+ },
304
+ "contact_sliding_speed": {
305
+ "max": 2.0296926498413086,
306
+ "mean": 0.5391670568348017,
307
+ "min": 0.004975942894816399,
308
+ "std": 0.5557337065044728
309
+ },
310
+ "foot_ground_error": {
311
+ "max": 0.012411188334226608,
312
+ "mean": 0.0062796306758469275,
313
+ "min": 0.0013083467492833734,
314
+ "std": 0.0025643484083098323
315
+ },
316
+ "num_samples": 43
317
+ },
318
+ "text": {
319
+ "generated": {
320
+ "matching_distance": {
321
+ "max": 1.310302972793579,
322
+ "mean": 1.2274910932363465,
323
+ "min": 1.134010910987854,
324
+ "std": 0.03699650409915577
325
+ },
326
+ "matching_score": 1.2274911403656006,
327
+ "mean_text_rank": 1.5,
328
+ "median_text_rank": 1.0,
329
+ "r_precision": [
330
+ 0.71875,
331
+ 0.90625,
332
+ 0.9375
333
+ ],
334
+ "retrieval_batch_size": 32,
335
+ "retrieval_num_batches": 1,
336
+ "retrieval_num_samples": 32,
337
+ "retrieval_order_policy": "manifest_order",
338
+ "retrieval_unique_datasets_per_batch": {
339
+ "max": 1.0,
340
+ "mean": 1.0,
341
+ "min": 1.0,
342
+ "std": 0.0
343
+ }
344
+ },
345
+ "reference": {
346
+ "matching_distance": {
347
+ "max": 1.3188464641571045,
348
+ "mean": 1.2265774760135384,
349
+ "min": 1.1549649238586426,
350
+ "std": 0.04007899902186925
351
+ },
352
+ "matching_score": 1.2265775203704834,
353
+ "mean_text_rank": 1.84375,
354
+ "median_text_rank": 1.0,
355
+ "r_precision": [
356
+ 0.6875,
357
+ 0.84375,
358
+ 0.90625
359
+ ],
360
+ "retrieval_batch_size": 32,
361
+ "retrieval_num_batches": 1,
362
+ "retrieval_num_samples": 32,
363
+ "retrieval_order_policy": "manifest_order",
364
+ "retrieval_unique_datasets_per_batch": {
365
+ "max": 1.0,
366
+ "mean": 1.0,
367
+ "min": 1.0,
368
+ "std": 0.0
369
+ }
370
+ }
371
+ },
372
+ "transition": {
373
+ "auj": {
374
+ "max": 1.9905637383757522,
375
+ "mean": 1.7810514533823214,
376
+ "min": 1.5876606825310595,
377
+ "std": 0.06310559905267207
378
+ },
379
+ "body_auj": {
380
+ "max": 1.9905637383757522,
381
+ "mean": 1.7810514533823214,
382
+ "min": 1.5876606825310595,
383
+ "std": 0.06310559905267207
384
+ },
385
+ "body_pj": {
386
+ "max": 0.12257324159145355,
387
+ "mean": 0.04556534072210969,
388
+ "min": 0.0018733292818069458,
389
+ "std": 0.03456947191402774
390
+ },
391
+ "boundary_frames": [
392
+ 60
393
+ ],
394
+ "chunk_length": 60,
395
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
396
+ "num_frames": 120,
397
+ "num_samples": 43,
398
+ "pj": {
399
+ "max": 0.12257324159145355,
400
+ "mean": 0.04556534072210969,
401
+ "min": 0.0018733292818069458,
402
+ "std": 0.03456947191402774
403
+ },
404
+ "qpos_auj": {
405
+ "max": 20.530484854076903,
406
+ "mean": 6.420968679559986,
407
+ "min": 4.301883842105374,
408
+ "std": 4.107536056696693
409
+ },
410
+ "qpos_pj": {
411
+ "max": 6.2786277532577515,
412
+ "mean": 0.8628149387530907,
413
+ "min": 0.019462376832962036,
414
+ "std": 1.4660741357554568
415
+ },
416
+ "reference_body_jerk_level": 0.032168975994990016,
417
+ "reference_qpos_jerk_level": 0.07551584170589375,
418
+ "transition_window_length": 60,
419
+ "transition_windows": [
420
+ [
421
+ 30,
422
+ 90
423
+ ]
424
+ ]
425
+ }
426
+ },
427
+ "fitness_original_filtered_test": {
428
+ "embedding": {
429
+ "diversity_generated": 1.1245486499453385,
430
+ "motion_fid": 1.3721207787485414,
431
+ "motion_kid": {
432
+ "mean": 0.0019768632581005363,
433
+ "std": 0.0
434
+ },
435
+ "num_samples": 43
436
+ },
437
+ "num_samples": 43,
438
+ "physical": {
439
+ "body_jerk_mean": {
440
+ "max": 307.0351257324219,
441
+ "mean": 62.3401137174562,
442
+ "min": 5.94058895111084,
443
+ "std": 74.72415873290508
444
+ },
445
+ "contact_sliding_speed": {
446
+ "max": 1.9512882232666016,
447
+ "mean": 0.20101934357351342,
448
+ "min": 0.007332059554755688,
449
+ "std": 0.4436527359467069
450
+ },
451
+ "foot_ground_error": {
452
+ "max": 0.012699918821454048,
453
+ "mean": 0.006037447639459441,
454
+ "min": 0.0020615984685719013,
455
+ "std": 0.0021288537281194384
456
+ },
457
+ "num_samples": 43
458
+ },
459
+ "text": {
460
+ "generated": {
461
+ "matching_distance": {
462
+ "max": 1.460925579071045,
463
+ "mean": 1.3796154177466105,
464
+ "min": 1.268312692642212,
465
+ "std": 0.04377553385148069
466
+ },
467
+ "matching_score": 1.3796154260635376,
468
+ "mean_text_rank": 10.46875,
469
+ "median_text_rank": 11.0,
470
+ "r_precision": [
471
+ 0.125,
472
+ 0.125,
473
+ 0.125
474
+ ],
475
+ "retrieval_batch_size": 32,
476
+ "retrieval_num_batches": 1,
477
+ "retrieval_num_samples": 32,
478
+ "retrieval_order_policy": "manifest_order",
479
+ "retrieval_unique_datasets_per_batch": {
480
+ "max": 1.0,
481
+ "mean": 1.0,
482
+ "min": 1.0,
483
+ "std": 0.0
484
+ }
485
+ },
486
+ "reference": {
487
+ "matching_distance": {
488
+ "max": 1.3819284439086914,
489
+ "mean": 1.2803652868714444,
490
+ "min": 1.209172010421753,
491
+ "std": 0.03319129105747331
492
+ },
493
+ "matching_score": 1.280365228652954,
494
+ "mean_text_rank": 7.5,
495
+ "median_text_rank": 7.5,
496
+ "r_precision": [
497
+ 0.21875,
498
+ 0.25,
499
+ 0.3125
500
+ ],
501
+ "retrieval_batch_size": 32,
502
+ "retrieval_num_batches": 1,
503
+ "retrieval_num_samples": 32,
504
+ "retrieval_order_policy": "manifest_order",
505
+ "retrieval_unique_datasets_per_batch": {
506
+ "max": 1.0,
507
+ "mean": 1.0,
508
+ "min": 1.0,
509
+ "std": 0.0
510
+ }
511
+ }
512
+ },
513
+ "transition": {
514
+ "auj": {
515
+ "max": 2.385969114632759,
516
+ "mean": 1.8030387962130037,
517
+ "min": 1.4960776350777631,
518
+ "std": 0.11265973690452995
519
+ },
520
+ "body_auj": {
521
+ "max": 2.385969114632759,
522
+ "mean": 1.8030387962130037,
523
+ "min": 1.4960776350777631,
524
+ "std": 0.11265973690452995
525
+ },
526
+ "body_pj": {
527
+ "max": 0.21083378558978438,
528
+ "mean": 0.04393309076468265,
529
+ "min": 0.0029095709323883057,
530
+ "std": 0.0441904998385772
531
+ },
532
+ "boundary_frames": [
533
+ 60
534
+ ],
535
+ "chunk_length": 60,
536
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
537
+ "num_frames": 120,
538
+ "num_samples": 43,
539
+ "pj": {
540
+ "max": 0.21083378558978438,
541
+ "mean": 0.04393309076468265,
542
+ "min": 0.0029095709323883057,
543
+ "std": 0.0441904998385772
544
+ },
545
+ "qpos_auj": {
546
+ "max": 144.49321365365512,
547
+ "mean": 7.995620672684568,
548
+ "min": 4.299943787984878,
549
+ "std": 21.08138489740843
550
+ },
551
+ "qpos_pj": {
552
+ "max": 12.581249475479126,
553
+ "mean": 0.4969454543586539,
554
+ "min": 0.011551141738891602,
555
+ "std": 1.8728602685043632
556
+ },
557
+ "reference_body_jerk_level": 0.032168975994990016,
558
+ "reference_qpos_jerk_level": 0.07551584170589375,
559
+ "transition_window_length": 60,
560
+ "transition_windows": [
561
+ [
562
+ 30,
563
+ 90
564
+ ]
565
+ ]
566
+ }
567
+ },
568
+ "humanml_original_filtered_test": {
569
+ "embedding": {
570
+ "diversity_generated": 1.1726888731015561,
571
+ "motion_fid": 1.175771434491134,
572
+ "motion_kid": {
573
+ "mean": 0.0014073873525246583,
574
+ "std": 0.0
575
+ },
576
+ "num_samples": 43
577
+ },
578
+ "num_samples": 43,
579
+ "physical": {
580
+ "body_jerk_mean": {
581
+ "max": 246.43716430664062,
582
+ "mean": 68.01800821548285,
583
+ "min": 2.1355345249176025,
584
+ "std": 57.236756620842854
585
+ },
586
+ "contact_sliding_speed": {
587
+ "max": 1.9895073175430298,
588
+ "mean": 0.39400377310812473,
589
+ "min": 0.0,
590
+ "std": 0.46918049539700873
591
+ },
592
+ "foot_ground_error": {
593
+ "max": 1.477441668510437,
594
+ "mean": 0.04231151855594024,
595
+ "min": 0.002110379049554467,
596
+ "std": 0.2215764948289639
597
+ },
598
+ "num_samples": 43
599
+ },
600
+ "text": {
601
+ "generated": {
602
+ "matching_distance": {
603
+ "max": 1.430628776550293,
604
+ "mean": 1.3250223120977713,
605
+ "min": 1.2387126684188843,
606
+ "std": 0.04734495346885312
607
+ },
608
+ "matching_score": 1.3250223398208618,
609
+ "mean_text_rank": 6.71875,
610
+ "median_text_rank": 3.0,
611
+ "r_precision": [
612
+ 0.28125,
613
+ 0.46875,
614
+ 0.5625
615
+ ],
616
+ "retrieval_batch_size": 32,
617
+ "retrieval_num_batches": 1,
618
+ "retrieval_num_samples": 32,
619
+ "retrieval_order_policy": "manifest_order",
620
+ "retrieval_unique_datasets_per_batch": {
621
+ "max": 1.0,
622
+ "mean": 1.0,
623
+ "min": 1.0,
624
+ "std": 0.0
625
+ }
626
+ },
627
+ "reference": {
628
+ "matching_distance": {
629
+ "max": 1.3049654960632324,
630
+ "mean": 1.2212754654329876,
631
+ "min": 1.151880145072937,
632
+ "std": 0.033672752576929775
633
+ },
634
+ "matching_score": 1.2212754487991333,
635
+ "mean_text_rank": 2.0625,
636
+ "median_text_rank": 1.0,
637
+ "r_precision": [
638
+ 0.53125,
639
+ 0.78125,
640
+ 0.875
641
+ ],
642
+ "retrieval_batch_size": 32,
643
+ "retrieval_num_batches": 1,
644
+ "retrieval_num_samples": 32,
645
+ "retrieval_order_policy": "manifest_order",
646
+ "retrieval_unique_datasets_per_batch": {
647
+ "max": 1.0,
648
+ "mean": 1.0,
649
+ "min": 1.0,
650
+ "std": 0.0
651
+ }
652
+ }
653
+ },
654
+ "transition": {
655
+ "auj": {
656
+ "max": 2.2740369624934447,
657
+ "mean": 1.7931604724291397,
658
+ "min": 1.6542664254946329,
659
+ "std": 0.08686737366358381
660
+ },
661
+ "body_auj": {
662
+ "max": 2.2740369624934447,
663
+ "mean": 1.7931604724291397,
664
+ "min": 1.6542664254946329,
665
+ "std": 0.08686737366358381
666
+ },
667
+ "body_pj": {
668
+ "max": 0.21533825621008873,
669
+ "mean": 0.04874728411046225,
670
+ "min": 0.0004047006368637085,
671
+ "std": 0.04271094174830256
672
+ },
673
+ "boundary_frames": [
674
+ 60
675
+ ],
676
+ "chunk_length": 60,
677
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
678
+ "num_frames": 120,
679
+ "num_samples": 43,
680
+ "pj": {
681
+ "max": 0.21533825621008873,
682
+ "mean": 0.04874728411046225,
683
+ "min": 0.0004047006368637085,
684
+ "std": 0.04271094174830256
685
+ },
686
+ "qpos_auj": {
687
+ "max": 10.246992179070007,
688
+ "mean": 5.0103711591434354,
689
+ "min": 4.300404030837447,
690
+ "std": 1.3281499654605524
691
+ },
692
+ "qpos_pj": {
693
+ "max": 2.85389244556427,
694
+ "mean": 0.3255547105367187,
695
+ "min": 0.0028911828994750977,
696
+ "std": 0.5239349095283596
697
+ },
698
+ "reference_body_jerk_level": 0.032168975994990016,
699
+ "reference_qpos_jerk_level": 0.07551584170589375,
700
+ "transition_window_length": 60,
701
+ "transition_windows": [
702
+ [
703
+ 30,
704
+ 90
705
+ ]
706
+ ]
707
+ }
708
+ },
709
+ "idea400_original_filtered_test": {
710
+ "embedding": {
711
+ "diversity_generated": 1.1694871863472327,
712
+ "motion_fid": 1.424413235305793,
713
+ "motion_kid": {
714
+ "mean": 0.003020059186301882,
715
+ "std": 0.0
716
+ },
717
+ "num_samples": 43
718
+ },
719
+ "num_samples": 43,
720
+ "physical": {
721
+ "body_jerk_mean": {
722
+ "max": 185.40188598632812,
723
+ "mean": 48.86681900467983,
724
+ "min": 2.1356678009033203,
725
+ "std": 52.217468326814576
726
+ },
727
+ "contact_sliding_speed": {
728
+ "max": 1.2346510887145996,
729
+ "mean": 0.27460066539365363,
730
+ "min": 0.0036743313539773226,
731
+ "std": 0.3540692147671962
732
+ },
733
+ "foot_ground_error": {
734
+ "max": 0.01912386529147625,
735
+ "mean": 0.0067057522327834095,
736
+ "min": 0.00249672238714993,
737
+ "std": 0.003613244239824124
738
+ },
739
+ "num_samples": 43
740
+ },
741
+ "text": {
742
+ "generated": {
743
+ "matching_distance": {
744
+ "max": 1.4311089515686035,
745
+ "mean": 1.3534246184105096,
746
+ "min": 1.2549099922180176,
747
+ "std": 0.040087949446493104
748
+ },
749
+ "matching_score": 1.3534245491027832,
750
+ "mean_text_rank": 8.5,
751
+ "median_text_rank": 5.0,
752
+ "r_precision": [
753
+ 0.1875,
754
+ 0.21875,
755
+ 0.3125
756
+ ],
757
+ "retrieval_batch_size": 32,
758
+ "retrieval_num_batches": 1,
759
+ "retrieval_num_samples": 32,
760
+ "retrieval_order_policy": "manifest_order",
761
+ "retrieval_unique_datasets_per_batch": {
762
+ "max": 1.0,
763
+ "mean": 1.0,
764
+ "min": 1.0,
765
+ "std": 0.0
766
+ }
767
+ },
768
+ "reference": {
769
+ "matching_distance": {
770
+ "max": 1.3933742046356201,
771
+ "mean": 1.253553493078365,
772
+ "min": 1.1590988636016846,
773
+ "std": 0.0395202842188391
774
+ },
775
+ "matching_score": 1.2535533905029297,
776
+ "mean_text_rank": 5.5625,
777
+ "median_text_rank": 4.0,
778
+ "r_precision": [
779
+ 0.21875,
780
+ 0.4375,
781
+ 0.46875
782
+ ],
783
+ "retrieval_batch_size": 32,
784
+ "retrieval_num_batches": 1,
785
+ "retrieval_num_samples": 32,
786
+ "retrieval_order_policy": "manifest_order",
787
+ "retrieval_unique_datasets_per_batch": {
788
+ "max": 1.0,
789
+ "mean": 1.0,
790
+ "min": 1.0,
791
+ "std": 0.0
792
+ }
793
+ }
794
+ },
795
+ "transition": {
796
+ "auj": {
797
+ "max": 1.9275732800142191,
798
+ "mean": 1.7972521730343467,
799
+ "min": 1.582865984803458,
800
+ "std": 0.06215447215088484
801
+ },
802
+ "body_auj": {
803
+ "max": 1.9275732800142191,
804
+ "mean": 1.7972521730343467,
805
+ "min": 1.582865984803458,
806
+ "std": 0.06215447215088484
807
+ },
808
+ "body_pj": {
809
+ "max": 0.19542989134788513,
810
+ "mean": 0.037036049071439474,
811
+ "min": 0.00043263286352157593,
812
+ "std": 0.04102339681952739
813
+ },
814
+ "boundary_frames": [
815
+ 60
816
+ ],
817
+ "chunk_length": 60,
818
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
819
+ "num_frames": 120,
820
+ "num_samples": 43,
821
+ "pj": {
822
+ "max": 0.19542989134788513,
823
+ "mean": 0.037036049071439474,
824
+ "min": 0.00043263286352157593,
825
+ "std": 0.04102339681952739
826
+ },
827
+ "qpos_auj": {
828
+ "max": 35.02763120426745,
829
+ "mean": 5.4219513973630304,
830
+ "min": 4.3006585100743475,
831
+ "std": 4.6576733588000225
832
+ },
833
+ "qpos_pj": {
834
+ "max": 12.468976974487305,
835
+ "mean": 0.5249752198849523,
836
+ "min": 0.002636820077896118,
837
+ "std": 1.8992311316861998
838
+ },
839
+ "reference_body_jerk_level": 0.032168975994990016,
840
+ "reference_qpos_jerk_level": 0.07551584170589375,
841
+ "transition_window_length": 60,
842
+ "transition_windows": [
843
+ [
844
+ 30,
845
+ 90
846
+ ]
847
+ ]
848
+ }
849
+ },
850
+ "lafan1_original_filtered_test": {
851
+ "embedding": {
852
+ "diversity_generated": 1.065131755766148,
853
+ "motion_fid": 0.9163817643240386,
854
+ "motion_kid": {
855
+ "mean": 0.001431817572137728,
856
+ "std": 0.0
857
+ },
858
+ "num_samples": 43
859
+ },
860
+ "num_samples": 43,
861
+ "physical": {
862
+ "body_jerk_mean": {
863
+ "max": 531.6823120117188,
864
+ "mean": 164.51493027043898,
865
+ "min": 6.81342077255249,
866
+ "std": 135.10997614115757
867
+ },
868
+ "contact_sliding_speed": {
869
+ "max": 2.914012908935547,
870
+ "mean": 0.8355859869354686,
871
+ "min": 0.008567381650209427,
872
+ "std": 0.7517330338413282
873
+ },
874
+ "foot_ground_error": {
875
+ "max": 0.013226441107690334,
876
+ "mean": 0.0067459214966051105,
877
+ "min": 0.0020911949686706066,
878
+ "std": 0.002395367031927562
879
+ },
880
+ "num_samples": 43
881
+ },
882
+ "text": {
883
+ "generated": {
884
+ "matching_distance": {
885
+ "max": 1.3756978511810303,
886
+ "mean": 1.2848902968473213,
887
+ "min": 1.22147798538208,
888
+ "std": 0.03506700337563933
889
+ },
890
+ "matching_score": 1.2848902940750122,
891
+ "mean_text_rank": 6.78125,
892
+ "median_text_rank": 5.0,
893
+ "r_precision": [
894
+ 0.375,
895
+ 0.40625,
896
+ 0.46875
897
+ ],
898
+ "retrieval_batch_size": 32,
899
+ "retrieval_num_batches": 1,
900
+ "retrieval_num_samples": 32,
901
+ "retrieval_order_policy": "manifest_order",
902
+ "retrieval_unique_datasets_per_batch": {
903
+ "max": 1.0,
904
+ "mean": 1.0,
905
+ "min": 1.0,
906
+ "std": 0.0
907
+ }
908
+ },
909
+ "reference": {
910
+ "matching_distance": {
911
+ "max": 1.3601514101028442,
912
+ "mean": 1.262607499610546,
913
+ "min": 1.1779136657714844,
914
+ "std": 0.036237216916877986
915
+ },
916
+ "matching_score": 1.262607455253601,
917
+ "mean_text_rank": 7.34375,
918
+ "median_text_rank": 4.0,
919
+ "r_precision": [
920
+ 0.28125,
921
+ 0.34375,
922
+ 0.46875
923
+ ],
924
+ "retrieval_batch_size": 32,
925
+ "retrieval_num_batches": 1,
926
+ "retrieval_num_samples": 32,
927
+ "retrieval_order_policy": "manifest_order",
928
+ "retrieval_unique_datasets_per_batch": {
929
+ "max": 1.0,
930
+ "mean": 1.0,
931
+ "min": 1.0,
932
+ "std": 0.0
933
+ }
934
+ }
935
+ },
936
+ "transition": {
937
+ "auj": {
938
+ "max": 4.683557439704222,
939
+ "mean": 1.9927798923981312,
940
+ "min": 1.7141404278200993,
941
+ "std": 0.5034205327935629
942
+ },
943
+ "body_auj": {
944
+ "max": 4.683557439704222,
945
+ "mean": 1.9927798923981312,
946
+ "min": 1.7141404278200993,
947
+ "std": 0.5034205327935629
948
+ },
949
+ "body_pj": {
950
+ "max": 0.8904187381267548,
951
+ "mean": 0.11073192904248487,
952
+ "min": 0.0009953081607818604,
953
+ "std": 0.13824281051787887
954
+ },
955
+ "boundary_frames": [
956
+ 60
957
+ ],
958
+ "chunk_length": 60,
959
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
960
+ "num_frames": 120,
961
+ "num_samples": 43,
962
+ "pj": {
963
+ "max": 0.8904187381267548,
964
+ "mean": 0.11073192904248487,
965
+ "min": 0.0009953081607818604,
966
+ "std": 0.13824281051787887
967
+ },
968
+ "qpos_auj": {
969
+ "max": 43.07620444184711,
970
+ "mean": 8.566258157005608,
971
+ "min": 4.297486459604036,
972
+ "std": 7.636360765758662
973
+ },
974
+ "qpos_pj": {
975
+ "max": 12.626699447631836,
976
+ "mean": 1.2561438795352398,
977
+ "min": 0.00521540641784668,
978
+ "std": 2.344095482079285
979
+ },
980
+ "reference_body_jerk_level": 0.032168975994990016,
981
+ "reference_qpos_jerk_level": 0.07551584170589375,
982
+ "transition_window_length": 60,
983
+ "transition_windows": [
984
+ [
985
+ 30,
986
+ 90
987
+ ]
988
+ ]
989
+ }
990
+ },
991
+ "motiongv_original_filtered_test": {
992
+ "embedding": {
993
+ "diversity_generated": 1.151175892575208,
994
+ "motion_fid": 1.4664699422882885,
995
+ "motion_kid": {
996
+ "mean": 0.0018693371237699097,
997
+ "std": 0.0
998
+ },
999
+ "num_samples": 43
1000
+ },
1001
+ "num_samples": 43,
1002
+ "physical": {
1003
+ "body_jerk_mean": {
1004
+ "max": 391.6789245605469,
1005
+ "mean": 82.85689859612043,
1006
+ "min": 3.0611748695373535,
1007
+ "std": 95.17589490486428
1008
+ },
1009
+ "contact_sliding_speed": {
1010
+ "max": 1.5828344821929932,
1011
+ "mean": 0.17748946919604097,
1012
+ "min": 0.006206035614013672,
1013
+ "std": 0.2848895833067525
1014
+ },
1015
+ "foot_ground_error": {
1016
+ "max": 0.05884654074907303,
1017
+ "mean": 0.00999552955266175,
1018
+ "min": 0.0020414358004927635,
1019
+ "std": 0.009597411811076375
1020
+ },
1021
+ "num_samples": 43
1022
+ },
1023
+ "text": {
1024
+ "generated": {
1025
+ "matching_distance": {
1026
+ "max": 1.472267508506775,
1027
+ "mean": 1.3855077416397805,
1028
+ "min": 1.2616697549819946,
1029
+ "std": 0.04268846702374364
1030
+ },
1031
+ "matching_score": 1.3855078220367432,
1032
+ "mean_text_rank": 6.71875,
1033
+ "median_text_rank": 3.5,
1034
+ "r_precision": [
1035
+ 0.28125,
1036
+ 0.40625,
1037
+ 0.5
1038
+ ],
1039
+ "retrieval_batch_size": 32,
1040
+ "retrieval_num_batches": 1,
1041
+ "retrieval_num_samples": 32,
1042
+ "retrieval_order_policy": "manifest_order",
1043
+ "retrieval_unique_datasets_per_batch": {
1044
+ "max": 1.0,
1045
+ "mean": 1.0,
1046
+ "min": 1.0,
1047
+ "std": 0.0
1048
+ }
1049
+ },
1050
+ "reference": {
1051
+ "matching_distance": {
1052
+ "max": 1.355556845664978,
1053
+ "mean": 1.2528588660927706,
1054
+ "min": 1.1668779850006104,
1055
+ "std": 0.04499687839541838
1056
+ },
1057
+ "matching_score": 1.2528588771820068,
1058
+ "mean_text_rank": 1.9375,
1059
+ "median_text_rank": 1.0,
1060
+ "r_precision": [
1061
+ 0.65625,
1062
+ 0.8125,
1063
+ 0.8125
1064
+ ],
1065
+ "retrieval_batch_size": 32,
1066
+ "retrieval_num_batches": 1,
1067
+ "retrieval_num_samples": 32,
1068
+ "retrieval_order_policy": "manifest_order",
1069
+ "retrieval_unique_datasets_per_batch": {
1070
+ "max": 1.0,
1071
+ "mean": 1.0,
1072
+ "min": 1.0,
1073
+ "std": 0.0
1074
+ }
1075
+ }
1076
+ },
1077
+ "transition": {
1078
+ "auj": {
1079
+ "max": 3.7732563196112654,
1080
+ "mean": 1.877907262608954,
1081
+ "min": 1.4860971376024386,
1082
+ "std": 0.3515215707606796
1083
+ },
1084
+ "body_auj": {
1085
+ "max": 3.7732563196112654,
1086
+ "mean": 1.877907262608954,
1087
+ "min": 1.4860971376024386,
1088
+ "std": 0.3515215707606796
1089
+ },
1090
+ "body_pj": {
1091
+ "max": 0.5607699304819107,
1092
+ "mean": 0.059313230828360416,
1093
+ "min": 0.0006179101765155792,
1094
+ "std": 0.09926593614052223
1095
+ },
1096
+ "boundary_frames": [
1097
+ 60
1098
+ ],
1099
+ "chunk_length": 60,
1100
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1101
+ "num_frames": 120,
1102
+ "num_samples": 43,
1103
+ "pj": {
1104
+ "max": 0.5607699304819107,
1105
+ "mean": 0.059313230828360416,
1106
+ "min": 0.0006179101765155792,
1107
+ "std": 0.09926593614052223
1108
+ },
1109
+ "qpos_auj": {
1110
+ "max": 62.14833038614482,
1111
+ "mean": 7.156893734841608,
1112
+ "min": 4.296862431809634,
1113
+ "std": 10.065915962574447
1114
+ },
1115
+ "qpos_pj": {
1116
+ "max": 12.888956546783447,
1117
+ "mean": 0.8218644026580245,
1118
+ "min": 0.003064531832933426,
1119
+ "std": 2.6184541571050173
1120
+ },
1121
+ "reference_body_jerk_level": 0.032168975994990016,
1122
+ "reference_qpos_jerk_level": 0.07551584170589375,
1123
+ "transition_window_length": 60,
1124
+ "transition_windows": [
1125
+ [
1126
+ 30,
1127
+ 90
1128
+ ]
1129
+ ]
1130
+ }
1131
+ },
1132
+ "motionllama_original_filtered_test": {
1133
+ "embedding": {
1134
+ "diversity_generated": 1.1930282606242615,
1135
+ "motion_fid": 1.3093164078052215,
1136
+ "motion_kid": {
1137
+ "mean": 0.001977269528855352,
1138
+ "std": 0.0
1139
+ },
1140
+ "num_samples": 42
1141
+ },
1142
+ "num_samples": 42,
1143
+ "physical": {
1144
+ "body_jerk_mean": {
1145
+ "max": 185.41676330566406,
1146
+ "mean": 17.548739166486833,
1147
+ "min": 2.4985158443450928,
1148
+ "std": 29.38218129714949
1149
+ },
1150
+ "contact_sliding_speed": {
1151
+ "max": 1.210018515586853,
1152
+ "mean": 0.05310293933421019,
1153
+ "min": 0.004471967928111553,
1154
+ "std": 0.18190039828460755
1155
+ },
1156
+ "foot_ground_error": {
1157
+ "max": 0.11203083395957947,
1158
+ "mean": 0.008613292068254114,
1159
+ "min": 0.0008450289606116712,
1160
+ "std": 0.016888853227158128
1161
+ },
1162
+ "num_samples": 42
1163
+ },
1164
+ "text": {
1165
+ "generated": {
1166
+ "matching_distance": {
1167
+ "max": 1.4584448337554932,
1168
+ "mean": 1.3611151774724324,
1169
+ "min": 1.2479358911514282,
1170
+ "std": 0.0521328721850203
1171
+ },
1172
+ "matching_score": 1.3611150979995728,
1173
+ "mean_text_rank": 7.75,
1174
+ "median_text_rank": 4.0,
1175
+ "r_precision": [
1176
+ 0.1875,
1177
+ 0.34375,
1178
+ 0.46875
1179
+ ],
1180
+ "retrieval_batch_size": 32,
1181
+ "retrieval_num_batches": 1,
1182
+ "retrieval_num_samples": 32,
1183
+ "retrieval_order_policy": "manifest_order",
1184
+ "retrieval_unique_datasets_per_batch": {
1185
+ "max": 1.0,
1186
+ "mean": 1.0,
1187
+ "min": 1.0,
1188
+ "std": 0.0
1189
+ }
1190
+ },
1191
+ "reference": {
1192
+ "matching_distance": {
1193
+ "max": 1.344225287437439,
1194
+ "mean": 1.2109532526561193,
1195
+ "min": 1.1528350114822388,
1196
+ "std": 0.03675615838449094
1197
+ },
1198
+ "matching_score": 1.2109531164169312,
1199
+ "mean_text_rank": 2.40625,
1200
+ "median_text_rank": 1.0,
1201
+ "r_precision": [
1202
+ 0.53125,
1203
+ 0.6875,
1204
+ 0.8125
1205
+ ],
1206
+ "retrieval_batch_size": 32,
1207
+ "retrieval_num_batches": 1,
1208
+ "retrieval_num_samples": 32,
1209
+ "retrieval_order_policy": "manifest_order",
1210
+ "retrieval_unique_datasets_per_batch": {
1211
+ "max": 1.0,
1212
+ "mean": 1.0,
1213
+ "min": 1.0,
1214
+ "std": 0.0
1215
+ }
1216
+ }
1217
+ },
1218
+ "transition": {
1219
+ "auj": {
1220
+ "max": 1.8308297701004421,
1221
+ "mean": 1.8203518604107993,
1222
+ "min": 1.7468128703884296,
1223
+ "std": 0.015218604769691082
1224
+ },
1225
+ "body_auj": {
1226
+ "max": 1.8308297701004421,
1227
+ "mean": 1.8203518604107993,
1228
+ "min": 1.7468128703884296,
1229
+ "std": 0.015218604769691082
1230
+ },
1231
+ "body_pj": {
1232
+ "max": 0.09689441323280334,
1233
+ "mean": 0.011014723008321155,
1234
+ "min": 0.0005841478705406189,
1235
+ "std": 0.017570536124353737
1236
+ },
1237
+ "boundary_frames": [
1238
+ 60
1239
+ ],
1240
+ "chunk_length": 60,
1241
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1242
+ "num_frames": 120,
1243
+ "num_samples": 42,
1244
+ "pj": {
1245
+ "max": 0.09689441323280334,
1246
+ "mean": 0.011014723008321155,
1247
+ "min": 0.0005841478705406189,
1248
+ "std": 0.017570536124353737
1249
+ },
1250
+ "qpos_auj": {
1251
+ "max": 9.52280969330879,
1252
+ "mean": 4.517935173208866,
1253
+ "min": 4.3012650981623235,
1254
+ "std": 0.8728498919105634
1255
+ },
1256
+ "qpos_pj": {
1257
+ "max": 2.8352200984954834,
1258
+ "mean": 0.1325104517579478,
1259
+ "min": 0.0033388733863830566,
1260
+ "std": 0.4389731704134287
1261
+ },
1262
+ "reference_body_jerk_level": 0.032168975994990016,
1263
+ "reference_qpos_jerk_level": 0.07551584170589375,
1264
+ "transition_window_length": 60,
1265
+ "transition_windows": [
1266
+ [
1267
+ 30,
1268
+ 90
1269
+ ]
1270
+ ]
1271
+ }
1272
+ },
1273
+ "omomo_original_filtered_test": {
1274
+ "embedding": {
1275
+ "diversity_generated": 1.1345490872611939,
1276
+ "motion_fid": 1.2325407801917518,
1277
+ "motion_kid": {
1278
+ "mean": 0.002298161219308703,
1279
+ "std": 0.0
1280
+ },
1281
+ "num_samples": 42
1282
+ },
1283
+ "num_samples": 42,
1284
+ "physical": {
1285
+ "body_jerk_mean": {
1286
+ "max": 215.9715118408203,
1287
+ "mean": 64.15830364681426,
1288
+ "min": 10.436027526855469,
1289
+ "std": 44.76208781970577
1290
+ },
1291
+ "contact_sliding_speed": {
1292
+ "max": 1.5735926628112793,
1293
+ "mean": 0.47663801983885823,
1294
+ "min": 0.015720246359705925,
1295
+ "std": 0.4178593234495149
1296
+ },
1297
+ "foot_ground_error": {
1298
+ "max": 0.01301527488976717,
1299
+ "mean": 0.006088023423217237,
1300
+ "min": 0.0020812139846384525,
1301
+ "std": 0.002377843957042252
1302
+ },
1303
+ "num_samples": 42
1304
+ },
1305
+ "text": {
1306
+ "generated": {
1307
+ "matching_distance": {
1308
+ "max": 1.433463454246521,
1309
+ "mean": 1.305754039968763,
1310
+ "min": 1.2346299886703491,
1311
+ "std": 0.042840413554061436
1312
+ },
1313
+ "matching_score": 1.3057539463043213,
1314
+ "mean_text_rank": 8.6875,
1315
+ "median_text_rank": 5.5,
1316
+ "r_precision": [
1317
+ 0.25,
1318
+ 0.3125,
1319
+ 0.40625
1320
+ ],
1321
+ "retrieval_batch_size": 32,
1322
+ "retrieval_num_batches": 1,
1323
+ "retrieval_num_samples": 32,
1324
+ "retrieval_order_policy": "manifest_order",
1325
+ "retrieval_unique_datasets_per_batch": {
1326
+ "max": 1.0,
1327
+ "mean": 1.0,
1328
+ "min": 1.0,
1329
+ "std": 0.0
1330
+ }
1331
+ },
1332
+ "reference": {
1333
+ "matching_distance": {
1334
+ "max": 1.3671505451202393,
1335
+ "mean": 1.269894642489297,
1336
+ "min": 1.2092769145965576,
1337
+ "std": 0.03030912014653846
1338
+ },
1339
+ "matching_score": 1.2698945999145508,
1340
+ "mean_text_rank": 8.40625,
1341
+ "median_text_rank": 6.5,
1342
+ "r_precision": [
1343
+ 0.15625,
1344
+ 0.21875,
1345
+ 0.25
1346
+ ],
1347
+ "retrieval_batch_size": 32,
1348
+ "retrieval_num_batches": 1,
1349
+ "retrieval_num_samples": 32,
1350
+ "retrieval_order_policy": "manifest_order",
1351
+ "retrieval_unique_datasets_per_batch": {
1352
+ "max": 1.0,
1353
+ "mean": 1.0,
1354
+ "min": 1.0,
1355
+ "std": 0.0
1356
+ }
1357
+ }
1358
+ },
1359
+ "transition": {
1360
+ "auj": {
1361
+ "max": 1.8620231315843276,
1362
+ "mean": 1.7725733755803437,
1363
+ "min": 1.6676089530947735,
1364
+ "std": 0.049378376046624706
1365
+ },
1366
+ "body_auj": {
1367
+ "max": 1.8620231315843276,
1368
+ "mean": 1.7725733755803437,
1369
+ "min": 1.6676089530947735,
1370
+ "std": 0.049378376046624706
1371
+ },
1372
+ "body_pj": {
1373
+ "max": 0.1809138059616089,
1374
+ "mean": 0.04543182488885664,
1375
+ "min": 0.004288120195269585,
1376
+ "std": 0.03565515372849116
1377
+ },
1378
+ "boundary_frames": [
1379
+ 60
1380
+ ],
1381
+ "chunk_length": 60,
1382
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1383
+ "num_frames": 120,
1384
+ "num_samples": 42,
1385
+ "pj": {
1386
+ "max": 0.1809138059616089,
1387
+ "mean": 0.04543182488885664,
1388
+ "min": 0.004288120195269585,
1389
+ "std": 0.03565515372849116
1390
+ },
1391
+ "qpos_auj": {
1392
+ "max": 28.471612076712674,
1393
+ "mean": 5.603556235023785,
1394
+ "min": 4.300355892869041,
1395
+ "std": 3.8283692334934805
1396
+ },
1397
+ "qpos_pj": {
1398
+ "max": 11.778656005859375,
1399
+ "mean": 0.6776692000927315,
1400
+ "min": 0.007280692458152771,
1401
+ "std": 1.8678563037401907
1402
+ },
1403
+ "reference_body_jerk_level": 0.032168975994990016,
1404
+ "reference_qpos_jerk_level": 0.07551584170589375,
1405
+ "transition_window_length": 60,
1406
+ "transition_windows": [
1407
+ [
1408
+ 30,
1409
+ 90
1410
+ ]
1411
+ ]
1412
+ }
1413
+ },
1414
+ "permo_original_filtered_test": {
1415
+ "embedding": {
1416
+ "diversity_generated": 1.0877195287239692,
1417
+ "motion_fid": 1.1566815401268762,
1418
+ "motion_kid": {
1419
+ "mean": 0.0022956662767517066,
1420
+ "std": 0.0
1421
+ },
1422
+ "num_samples": 42
1423
+ },
1424
+ "num_samples": 42,
1425
+ "physical": {
1426
+ "body_jerk_mean": {
1427
+ "max": 263.7925720214844,
1428
+ "mean": 76.4998429048629,
1429
+ "min": 4.8154168128967285,
1430
+ "std": 69.10691261564781
1431
+ },
1432
+ "contact_sliding_speed": {
1433
+ "max": 1.7778644561767578,
1434
+ "mean": 0.31765077484860305,
1435
+ "min": 0.009715426713228226,
1436
+ "std": 0.403836893708808
1437
+ },
1438
+ "foot_ground_error": {
1439
+ "max": 0.025011582300066948,
1440
+ "mean": 0.007566677107076559,
1441
+ "min": 0.0026211035437881947,
1442
+ "std": 0.004112016751390115
1443
+ },
1444
+ "num_samples": 42
1445
+ },
1446
+ "text": {
1447
+ "generated": {
1448
+ "matching_distance": {
1449
+ "max": 1.3902548551559448,
1450
+ "mean": 1.3313227948688326,
1451
+ "min": 1.2616204023361206,
1452
+ "std": 0.0340989370710566
1453
+ },
1454
+ "matching_score": 1.3313229084014893,
1455
+ "mean_text_rank": 4.84375,
1456
+ "median_text_rank": 4.0,
1457
+ "r_precision": [
1458
+ 0.375,
1459
+ 0.40625,
1460
+ 0.4375
1461
+ ],
1462
+ "retrieval_batch_size": 32,
1463
+ "retrieval_num_batches": 1,
1464
+ "retrieval_num_samples": 32,
1465
+ "retrieval_order_policy": "manifest_order",
1466
+ "retrieval_unique_datasets_per_batch": {
1467
+ "max": 1.0,
1468
+ "mean": 1.0,
1469
+ "min": 1.0,
1470
+ "std": 0.0
1471
+ }
1472
+ },
1473
+ "reference": {
1474
+ "matching_distance": {
1475
+ "max": 1.3314403295516968,
1476
+ "mean": 1.2064374969119118,
1477
+ "min": 1.1483595371246338,
1478
+ "std": 0.03169944929872626
1479
+ },
1480
+ "matching_score": 1.2064374685287476,
1481
+ "mean_text_rank": 4.34375,
1482
+ "median_text_rank": 3.0,
1483
+ "r_precision": [
1484
+ 0.25,
1485
+ 0.4375,
1486
+ 0.53125
1487
+ ],
1488
+ "retrieval_batch_size": 32,
1489
+ "retrieval_num_batches": 1,
1490
+ "retrieval_num_samples": 32,
1491
+ "retrieval_order_policy": "manifest_order",
1492
+ "retrieval_unique_datasets_per_batch": {
1493
+ "max": 1.0,
1494
+ "mean": 1.0,
1495
+ "min": 1.0,
1496
+ "std": 0.0
1497
+ }
1498
+ }
1499
+ },
1500
+ "transition": {
1501
+ "auj": {
1502
+ "max": 2.781857592675066,
1503
+ "mean": 1.8609303950500373,
1504
+ "min": 1.5932010662360152,
1505
+ "std": 0.2132839801381445
1506
+ },
1507
+ "body_auj": {
1508
+ "max": 2.781857592675066,
1509
+ "mean": 1.8609303950500373,
1510
+ "min": 1.5932010662360152,
1511
+ "std": 0.2132839801381445
1512
+ },
1513
+ "body_pj": {
1514
+ "max": 0.2385309636592865,
1515
+ "mean": 0.0728943720903425,
1516
+ "min": 0.0014599980786442757,
1517
+ "std": 0.0661067180136749
1518
+ },
1519
+ "boundary_frames": [
1520
+ 60
1521
+ ],
1522
+ "chunk_length": 60,
1523
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1524
+ "num_frames": 120,
1525
+ "num_samples": 42,
1526
+ "pj": {
1527
+ "max": 0.2385309636592865,
1528
+ "mean": 0.0728943720903425,
1529
+ "min": 0.0014599980786442757,
1530
+ "std": 0.0661067180136749
1531
+ },
1532
+ "qpos_auj": {
1533
+ "max": 12.99549435606882,
1534
+ "mean": 5.456390068404779,
1535
+ "min": 4.300785458770909,
1536
+ "std": 1.9640080094470993
1537
+ },
1538
+ "qpos_pj": {
1539
+ "max": 3.9514918252825737,
1540
+ "mean": 0.4679690366056526,
1541
+ "min": 0.006184376776218414,
1542
+ "std": 0.7646300858697805
1543
+ },
1544
+ "reference_body_jerk_level": 0.032168975994990016,
1545
+ "reference_qpos_jerk_level": 0.07551584170589375,
1546
+ "transition_window_length": 60,
1547
+ "transition_windows": [
1548
+ [
1549
+ 30,
1550
+ 90
1551
+ ]
1552
+ ]
1553
+ }
1554
+ },
1555
+ "snapmogen_original_filtered_test": {
1556
+ "embedding": {
1557
+ "diversity_generated": 1.2069637839684089,
1558
+ "motion_fid": 1.3497101089407968,
1559
+ "motion_kid": {
1560
+ "mean": 0.0020190263621038262,
1561
+ "std": 0.0
1562
+ },
1563
+ "num_samples": 42
1564
+ },
1565
+ "num_samples": 42,
1566
+ "physical": {
1567
+ "body_jerk_mean": {
1568
+ "max": 228.76815795898438,
1569
+ "mean": 39.739420181229,
1570
+ "min": 2.503999710083008,
1571
+ "std": 55.302429574414276
1572
+ },
1573
+ "contact_sliding_speed": {
1574
+ "max": 1.628731369972229,
1575
+ "mean": 0.2116111345655684,
1576
+ "min": 0.006765882950276136,
1577
+ "std": 0.36438021456289116
1578
+ },
1579
+ "foot_ground_error": {
1580
+ "max": 0.011111666448414326,
1581
+ "mean": 0.005569282187415021,
1582
+ "min": 0.0005915897199884057,
1583
+ "std": 0.0024525098901194337
1584
+ },
1585
+ "num_samples": 42
1586
+ },
1587
+ "text": {
1588
+ "generated": {
1589
+ "matching_distance": {
1590
+ "max": 1.4769279956817627,
1591
+ "mean": 1.377476453781128,
1592
+ "min": 1.2130833864212036,
1593
+ "std": 0.046379623782035695
1594
+ },
1595
+ "matching_score": 1.377476453781128,
1596
+ "mean_text_rank": 6.84375,
1597
+ "median_text_rank": 4.0,
1598
+ "r_precision": [
1599
+ 0.34375,
1600
+ 0.4375,
1601
+ 0.46875
1602
+ ],
1603
+ "retrieval_batch_size": 32,
1604
+ "retrieval_num_batches": 1,
1605
+ "retrieval_num_samples": 32,
1606
+ "retrieval_order_policy": "manifest_order",
1607
+ "retrieval_unique_datasets_per_batch": {
1608
+ "max": 1.0,
1609
+ "mean": 1.0,
1610
+ "min": 1.0,
1611
+ "std": 0.0
1612
+ }
1613
+ },
1614
+ "reference": {
1615
+ "matching_distance": {
1616
+ "max": 1.2890117168426514,
1617
+ "mean": 1.2102941047577631,
1618
+ "min": 1.1285470724105835,
1619
+ "std": 0.04065653924035539
1620
+ },
1621
+ "matching_score": 1.2102941274642944,
1622
+ "mean_text_rank": 1.75,
1623
+ "median_text_rank": 1.0,
1624
+ "r_precision": [
1625
+ 0.75,
1626
+ 0.8125,
1627
+ 0.84375
1628
+ ],
1629
+ "retrieval_batch_size": 32,
1630
+ "retrieval_num_batches": 1,
1631
+ "retrieval_num_samples": 32,
1632
+ "retrieval_order_policy": "manifest_order",
1633
+ "retrieval_unique_datasets_per_batch": {
1634
+ "max": 1.0,
1635
+ "mean": 1.0,
1636
+ "min": 1.0,
1637
+ "std": 0.0
1638
+ }
1639
+ }
1640
+ },
1641
+ "transition": {
1642
+ "auj": {
1643
+ "max": 5.563047105616452,
1644
+ "mean": 1.9004562439177437,
1645
+ "min": 1.720805086064156,
1646
+ "std": 0.5739356354454959
1647
+ },
1648
+ "body_auj": {
1649
+ "max": 5.563047105616452,
1650
+ "mean": 1.9004562439177437,
1651
+ "min": 1.720805086064156,
1652
+ "std": 0.5739356354454959
1653
+ },
1654
+ "body_pj": {
1655
+ "max": 1.3313361704349518,
1656
+ "mean": 0.058183968610440694,
1657
+ "min": 0.0004454776644706726,
1658
+ "std": 0.20293248270696373
1659
+ },
1660
+ "boundary_frames": [
1661
+ 60
1662
+ ],
1663
+ "chunk_length": 60,
1664
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1665
+ "num_frames": 120,
1666
+ "num_samples": 42,
1667
+ "pj": {
1668
+ "max": 1.3313361704349518,
1669
+ "mean": 0.058183968610440694,
1670
+ "min": 0.0004454776644706726,
1671
+ "std": 0.20293248270696373
1672
+ },
1673
+ "qpos_auj": {
1674
+ "max": 52.49802010814777,
1675
+ "mean": 6.607324879388726,
1676
+ "min": 4.299905554166392,
1677
+ "std": 8.780425573903713
1678
+ },
1679
+ "qpos_pj": {
1680
+ "max": 12.550551891326904,
1681
+ "mean": 0.81523050702942,
1682
+ "min": 0.0026228129863739014,
1683
+ "std": 2.5778904370957694
1684
+ },
1685
+ "reference_body_jerk_level": 0.032168975994990016,
1686
+ "reference_qpos_jerk_level": 0.07551584170589375,
1687
+ "transition_window_length": 60,
1688
+ "transition_windows": [
1689
+ [
1690
+ 30,
1691
+ 90
1692
+ ]
1693
+ ]
1694
+ }
1695
+ }
1696
+ },
1697
+ "embedding_frames": 120,
1698
+ "embedding_metrics": {
1699
+ "diversity_generated": 1.185870855502696,
1700
+ "embedding_frames": 120,
1701
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_040000.pt",
1702
+ "matching_score_generated": 1.3311593532562256,
1703
+ "matching_score_reference": 1.244696855545044,
1704
+ "motion_fid": 0.5886926257750928,
1705
+ "motion_key": "body_pos_local",
1706
+ "motion_kid": {
1707
+ "mean": 0.0009367047902468784,
1708
+ "std": 0.0
1709
+ },
1710
+ "multimodality": null,
1711
+ "num_samples": 512,
1712
+ "r_precision_generated": [
1713
+ 0.240234375,
1714
+ 0.33984375,
1715
+ 0.412109375
1716
+ ],
1717
+ "r_precision_reference": [
1718
+ 0.705078125,
1719
+ 0.8359375,
1720
+ 0.90234375
1721
+ ],
1722
+ "text_cache_dir": "outputs/evaluator_text_cache/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f",
1723
+ "text_encoder_model": "${OMG_MODELS_ROOT}/text_encoders/t5-3b-local",
1724
+ "text_generated": {
1725
+ "matching_distance": {
1726
+ "max": 1.4769279956817627,
1727
+ "mean": 1.331159382360056,
1728
+ "min": 1.134010910987854,
1729
+ "std": 0.06046604617699041
1730
+ },
1731
+ "matching_score": 1.3311593532562256,
1732
+ "mean_text_rank": 8.705078125,
1733
+ "median_text_rank": 6.0,
1734
+ "r_precision": [
1735
+ 0.240234375,
1736
+ 0.33984375,
1737
+ 0.412109375
1738
+ ],
1739
+ "retrieval_batch_size": 32,
1740
+ "retrieval_num_batches": 16,
1741
+ "retrieval_num_samples": 512,
1742
+ "retrieval_order_policy": "stratified_by_dataset",
1743
+ "retrieval_unique_datasets_per_batch": {
1744
+ "max": 12.0,
1745
+ "mean": 12.0,
1746
+ "min": 12.0,
1747
+ "std": 0.0
1748
+ }
1749
+ },
1750
+ "text_reference": {
1751
+ "matching_distance": {
1752
+ "max": 1.4135501384735107,
1753
+ "mean": 1.2446968206204474,
1754
+ "min": 1.1285470724105835,
1755
+ "std": 0.0457579048836768
1756
+ },
1757
+ "matching_score": 1.244696855545044,
1758
+ "mean_text_rank": 1.869140625,
1759
+ "median_text_rank": 1.0,
1760
+ "r_precision": [
1761
+ 0.705078125,
1762
+ 0.8359375,
1763
+ 0.90234375
1764
+ ],
1765
+ "retrieval_batch_size": 32,
1766
+ "retrieval_num_batches": 16,
1767
+ "retrieval_num_samples": 512,
1768
+ "retrieval_order_policy": "stratified_by_dataset",
1769
+ "retrieval_unique_datasets_per_batch": {
1770
+ "max": 12.0,
1771
+ "mean": 12.0,
1772
+ "min": 12.0,
1773
+ "std": 0.0
1774
+ }
1775
+ },
1776
+ "text_retrieval_batch_size": 32
1777
+ },
1778
+ "exp": "g1_transformer_guided_300m",
1779
+ "generated_qpos": "${OMG_REPO_ROOT}/outputs_baseline/kimodo/g1_seed/text_mixed_modalities_all_v1_120f_transition/generated_qpos.npz",
1780
+ "label": "Kimodo-G1-SEED-v1",
1781
+ "num_frames": 120,
1782
+ "num_samples": 512,
1783
+ "physical_metrics": {
1784
+ "body_jerk_mean": {
1785
+ "max": 531.6823120117188,
1786
+ "mean": 72.15493237087503,
1787
+ "min": 2.1355345249176025,
1788
+ "std": 79.05384587679872
1789
+ },
1790
+ "contact_height_threshold": 0.12,
1791
+ "contact_penetration_tolerance": 0.02,
1792
+ "contact_sliding_speed": {
1793
+ "max": 2.914012908935547,
1794
+ "mean": 0.3516153497353116,
1795
+ "min": 0.0,
1796
+ "std": 0.475502377479124
1797
+ },
1798
+ "foot_ground_error": {
1799
+ "max": 1.477441668510437,
1800
+ "mean": 0.009914551770975777,
1801
+ "min": 0.0005915897199884057,
1802
+ "std": 0.06528683401820799
1803
+ },
1804
+ "frames": 120,
1805
+ "num_samples": 512
1806
+ },
1807
+ "real_motion_metrics": {
1808
+ "diversity_real": 1.322922111091206,
1809
+ "matching_distance_real": {
1810
+ "max": 1.4135501384735107,
1811
+ "mean": 1.2446968206204474,
1812
+ "min": 1.1285470724105835,
1813
+ "std": 0.0457579048836768
1814
+ },
1815
+ "matching_score_real": 1.244696855545044,
1816
+ "motion_fid_reference_eval": 0.0,
1817
+ "motion_fid_split_half": 0.0,
1818
+ "motion_kid_reference_eval": {
1819
+ "mean": -1.9876635613869098e-05,
1820
+ "std": 0.0
1821
+ },
1822
+ "motion_kid_split_half": {
1823
+ "mean": -1.9876635613869098e-05,
1824
+ "std": 0.0
1825
+ },
1826
+ "multimodality": "-",
1827
+ "num_reference_samples": 512,
1828
+ "num_samples": 512,
1829
+ "r_precision_real": [
1830
+ 0.455078125,
1831
+ 0.587890625,
1832
+ 0.677734375
1833
+ ],
1834
+ "text_real": {
1835
+ "matching_distance": {
1836
+ "max": 1.4135501384735107,
1837
+ "mean": 1.2446968206204474,
1838
+ "min": 1.1285470724105835,
1839
+ "std": 0.0457579048836768
1840
+ },
1841
+ "matching_score": 1.244696855545044,
1842
+ "mean_text_rank": 3.904296875,
1843
+ "median_text_rank": 2.0,
1844
+ "r_precision": [
1845
+ 0.455078125,
1846
+ 0.587890625,
1847
+ 0.677734375
1848
+ ],
1849
+ "retrieval_batch_size": 32,
1850
+ "retrieval_num_batches": 16,
1851
+ "retrieval_num_samples": 512,
1852
+ "retrieval_order_policy": "manifest_order",
1853
+ "retrieval_unique_datasets_per_batch": null
1854
+ }
1855
+ },
1856
+ "reference_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/reference_qpos.npz",
1857
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/sample_metrics.jsonl",
1858
+ "sample_rankings_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/sample_rankings.json",
1859
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_samples/mixed_modalities_all_v1/text_test_512_120f.jsonl",
1860
+ "split": "test",
1861
+ "tracker_executed": {
1862
+ "enabled": false
1863
+ },
1864
+ "transition_metrics": {
1865
+ "auj": {
1866
+ "max": 6.514222457101631,
1867
+ "mean": 1.8406120086684412,
1868
+ "min": 1.4860971376024386,
1869
+ "std": 0.3403637125100069
1870
+ },
1871
+ "body_auj": {
1872
+ "max": 6.514222457101631,
1873
+ "mean": 1.8406120086684412,
1874
+ "min": 1.4860971376024386,
1875
+ "std": 0.3403637125100069
1876
+ },
1877
+ "body_pj": {
1878
+ "max": 1.9185047969222069,
1879
+ "mean": 0.0578470742943864,
1880
+ "min": 0.0004047006368637085,
1881
+ "std": 0.12257751816949211
1882
+ },
1883
+ "boundary_frames": [
1884
+ 60
1885
+ ],
1886
+ "chunk_length": 60,
1887
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1888
+ "enabled": true,
1889
+ "num_frames": 120,
1890
+ "num_samples": 512,
1891
+ "pj": {
1892
+ "max": 1.9185047969222069,
1893
+ "mean": 0.0578470742943864,
1894
+ "min": 0.0004047006368637085,
1895
+ "std": 0.12257751816949211
1896
+ },
1897
+ "qpos_auj": {
1898
+ "max": 144.49321365365512,
1899
+ "mean": 6.459746310243598,
1900
+ "min": 4.2940439462623115,
1901
+ "std": 8.748215194373627
1902
+ },
1903
+ "qpos_pj": {
1904
+ "max": 12.888956546783447,
1905
+ "mean": 0.706420054541411,
1906
+ "min": 0.0026228129863739014,
1907
+ "std": 1.8797036959973208
1908
+ },
1909
+ "reference_body_jerk_level": 0.032168975994990016,
1910
+ "reference_qpos_jerk_level": 0.07551584170589375,
1911
+ "transition_window_length": 60,
1912
+ "transition_windows": [
1913
+ [
1914
+ 30,
1915
+ 90
1916
+ ]
1917
+ ]
1918
+ }
1919
+ }
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/dataset_metrics.json ADDED
@@ -0,0 +1,1694 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "100style_original_filtered_test": {
3
+ "embedding": {
4
+ "diversity_generated": 1.0704150905599346,
5
+ "motion_fid": 1.2838937050444919,
6
+ "motion_kid": {
7
+ "mean": 0.0022249471798367892,
8
+ "std": 0.0
9
+ },
10
+ "num_samples": 43
11
+ },
12
+ "num_samples": 43,
13
+ "physical": {
14
+ "body_jerk_mean": {
15
+ "max": 154.36744689941406,
16
+ "mean": 73.70867042208828,
17
+ "min": 2.9786179065704346,
18
+ "std": 37.651289022141334
19
+ },
20
+ "contact_sliding_speed": {
21
+ "max": 1.0541962385177612,
22
+ "mean": 0.3344618064286404,
23
+ "min": 0.007166638039052486,
24
+ "std": 0.2880695016780662
25
+ },
26
+ "foot_ground_error": {
27
+ "max": 0.015996823087334633,
28
+ "mean": 0.005727387244574899,
29
+ "min": 0.0019035614095628262,
30
+ "std": 0.002544252817203053
31
+ },
32
+ "num_samples": 43
33
+ },
34
+ "text": {
35
+ "generated": {
36
+ "matching_distance": {
37
+ "max": 1.4084994792938232,
38
+ "mean": 1.3151017427444458,
39
+ "min": 1.2379783391952515,
40
+ "std": 0.036907895016914875
41
+ },
42
+ "matching_score": 1.3151017427444458,
43
+ "mean_text_rank": 10.90625,
44
+ "median_text_rank": 10.0,
45
+ "r_precision": [
46
+ 0.09375,
47
+ 0.125,
48
+ 0.1875
49
+ ],
50
+ "retrieval_batch_size": 32,
51
+ "retrieval_num_batches": 1,
52
+ "retrieval_num_samples": 32,
53
+ "retrieval_order_policy": "manifest_order",
54
+ "retrieval_unique_datasets_per_batch": {
55
+ "max": 1.0,
56
+ "mean": 1.0,
57
+ "min": 1.0,
58
+ "std": 0.0
59
+ }
60
+ },
61
+ "reference": {
62
+ "matching_distance": {
63
+ "max": 1.3596904277801514,
64
+ "mean": 1.26676583567331,
65
+ "min": 1.1853543519973755,
66
+ "std": 0.03781489479220778
67
+ },
68
+ "matching_score": 1.266765832901001,
69
+ "mean_text_rank": 5.71875,
70
+ "median_text_rank": 3.5,
71
+ "r_precision": [
72
+ 0.34375,
73
+ 0.46875,
74
+ 0.5
75
+ ],
76
+ "retrieval_batch_size": 32,
77
+ "retrieval_num_batches": 1,
78
+ "retrieval_num_samples": 32,
79
+ "retrieval_order_policy": "manifest_order",
80
+ "retrieval_unique_datasets_per_batch": {
81
+ "max": 1.0,
82
+ "mean": 1.0,
83
+ "min": 1.0,
84
+ "std": 0.0
85
+ }
86
+ }
87
+ },
88
+ "transition": {
89
+ "auj": {
90
+ "max": 3.0685779272303186,
91
+ "mean": 1.8053734862221296,
92
+ "min": 1.6700049626518585,
93
+ "std": 0.20044042082439859
94
+ },
95
+ "body_auj": {
96
+ "max": 3.0685779272303186,
97
+ "mean": 1.8053734862221296,
98
+ "min": 1.6700049626518585,
99
+ "std": 0.20044042082439859
100
+ },
101
+ "body_pj": {
102
+ "max": 0.6321938335895538,
103
+ "mean": 0.06189292100756321,
104
+ "min": 0.0008932948112487793,
105
+ "std": 0.09465196402988692
106
+ },
107
+ "boundary_frames": [
108
+ 60
109
+ ],
110
+ "chunk_length": 60,
111
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
112
+ "num_frames": 120,
113
+ "num_samples": 43,
114
+ "pj": {
115
+ "max": 0.6321938335895538,
116
+ "mean": 0.06189292100756321,
117
+ "min": 0.0008932948112487793,
118
+ "std": 0.09465196402988692
119
+ },
120
+ "qpos_auj": {
121
+ "max": 24.71011780718753,
122
+ "mean": 6.1841539223100535,
123
+ "min": 4.2977976540647,
124
+ "std": 3.838152676323723
125
+ },
126
+ "qpos_pj": {
127
+ "max": 9.507055640220642,
128
+ "mean": 0.9185523066323164,
129
+ "min": 0.003028661012649536,
130
+ "std": 1.6940965662424199
131
+ },
132
+ "reference_body_jerk_level": 0.032168975994990016,
133
+ "reference_qpos_jerk_level": 0.07551584170589375,
134
+ "transition_window_length": 60,
135
+ "transition_windows": [
136
+ [
137
+ 30,
138
+ 90
139
+ ]
140
+ ]
141
+ }
142
+ },
143
+ "amass_original_filtered_test": {
144
+ "embedding": {
145
+ "diversity_generated": 1.111925104779673,
146
+ "motion_fid": 1.328561012147938,
147
+ "motion_kid": {
148
+ "mean": 0.0022260718776268007,
149
+ "std": 0.0
150
+ },
151
+ "num_samples": 43
152
+ },
153
+ "num_samples": 43,
154
+ "physical": {
155
+ "body_jerk_mean": {
156
+ "max": 269.9172668457031,
157
+ "mean": 95.9706466918768,
158
+ "min": 2.5841948986053467,
159
+ "std": 70.75177739661642
160
+ },
161
+ "contact_sliding_speed": {
162
+ "max": 1.5415642261505127,
163
+ "mean": 0.39597279582779077,
164
+ "min": 0.005511525087058544,
165
+ "std": 0.40617249845569314
166
+ },
167
+ "foot_ground_error": {
168
+ "max": 0.04935852810740471,
169
+ "mean": 0.00705925366559694,
170
+ "min": 0.0024636401794850826,
171
+ "std": 0.007562361206799781
172
+ },
173
+ "num_samples": 43
174
+ },
175
+ "text": {
176
+ "generated": {
177
+ "matching_distance": {
178
+ "max": 1.418434739112854,
179
+ "mean": 1.3283776654753574,
180
+ "min": 1.2401793003082275,
181
+ "std": 0.04136207243662587
182
+ },
183
+ "matching_score": 1.328377604484558,
184
+ "mean_text_rank": 6.96875,
185
+ "median_text_rank": 3.0,
186
+ "r_precision": [
187
+ 0.375,
188
+ 0.5,
189
+ 0.5
190
+ ],
191
+ "retrieval_batch_size": 32,
192
+ "retrieval_num_batches": 1,
193
+ "retrieval_num_samples": 32,
194
+ "retrieval_order_policy": "manifest_order",
195
+ "retrieval_unique_datasets_per_batch": {
196
+ "max": 1.0,
197
+ "mean": 1.0,
198
+ "min": 1.0,
199
+ "std": 0.0
200
+ }
201
+ },
202
+ "reference": {
203
+ "matching_distance": {
204
+ "max": 1.4135501384735107,
205
+ "mean": 1.272889874702276,
206
+ "min": 1.172876000404358,
207
+ "std": 0.04160945025899305
208
+ },
209
+ "matching_score": 1.2728898525238037,
210
+ "mean_text_rank": 4.40625,
211
+ "median_text_rank": 2.0,
212
+ "r_precision": [
213
+ 0.3125,
214
+ 0.5625,
215
+ 0.625
216
+ ],
217
+ "retrieval_batch_size": 32,
218
+ "retrieval_num_batches": 1,
219
+ "retrieval_num_samples": 32,
220
+ "retrieval_order_policy": "manifest_order",
221
+ "retrieval_unique_datasets_per_batch": {
222
+ "max": 1.0,
223
+ "mean": 1.0,
224
+ "min": 1.0,
225
+ "std": 0.0
226
+ }
227
+ }
228
+ },
229
+ "transition": {
230
+ "auj": {
231
+ "max": 6.514222457101631,
232
+ "mean": 1.8822794797577118,
233
+ "min": 1.5973546691528981,
234
+ "std": 0.7210500357139227
235
+ },
236
+ "body_auj": {
237
+ "max": 6.514222457101631,
238
+ "mean": 1.8822794797577118,
239
+ "min": 1.5973546691528981,
240
+ "std": 0.7210500357139227
241
+ },
242
+ "body_pj": {
243
+ "max": 1.9185047969222069,
244
+ "mean": 0.09840007811828061,
245
+ "min": 0.00058785080909729,
246
+ "std": 0.28364867706203306
247
+ },
248
+ "boundary_frames": [
249
+ 60
250
+ ],
251
+ "chunk_length": 60,
252
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
253
+ "num_frames": 120,
254
+ "num_samples": 43,
255
+ "pj": {
256
+ "max": 1.9185047969222069,
257
+ "mean": 0.09840007811828061,
258
+ "min": 0.00058785080909729,
259
+ "std": 0.28364867706203306
260
+ },
261
+ "qpos_auj": {
262
+ "max": 72.83266460476783,
263
+ "mean": 8.4905600420132,
264
+ "min": 4.2940439462623115,
265
+ "std": 11.744696961776553
266
+ },
267
+ "qpos_pj": {
268
+ "max": 12.584792256355286,
269
+ "mean": 1.159780290310791,
270
+ "min": 0.002947501838207245,
271
+ "std": 2.334767403530377
272
+ },
273
+ "reference_body_jerk_level": 0.032168975994990016,
274
+ "reference_qpos_jerk_level": 0.07551584170589375,
275
+ "transition_window_length": 60,
276
+ "transition_windows": [
277
+ [
278
+ 30,
279
+ 90
280
+ ]
281
+ ]
282
+ }
283
+ },
284
+ "bones_seed_original_filtered_test": {
285
+ "embedding": {
286
+ "diversity_generated": 1.1937374883617156,
287
+ "motion_fid": 0.4045952532270245,
288
+ "motion_kid": {
289
+ "mean": -0.00010609229304492374,
290
+ "std": 0.0
291
+ },
292
+ "num_samples": 43
293
+ },
294
+ "num_samples": 43,
295
+ "physical": {
296
+ "body_jerk_mean": {
297
+ "max": 261.4084777832031,
298
+ "mean": 69.5281113635662,
299
+ "min": 3.05818772315979,
300
+ "std": 68.85240862628845
301
+ },
302
+ "contact_sliding_speed": {
303
+ "max": 2.0296926498413086,
304
+ "mean": 0.5391670568348017,
305
+ "min": 0.004975942894816399,
306
+ "std": 0.5557337065044728
307
+ },
308
+ "foot_ground_error": {
309
+ "max": 0.012411188334226608,
310
+ "mean": 0.0062796306758469275,
311
+ "min": 0.0013083467492833734,
312
+ "std": 0.0025643484083098323
313
+ },
314
+ "num_samples": 43
315
+ },
316
+ "text": {
317
+ "generated": {
318
+ "matching_distance": {
319
+ "max": 1.310302972793579,
320
+ "mean": 1.2274910932363465,
321
+ "min": 1.134010910987854,
322
+ "std": 0.03699650409915577
323
+ },
324
+ "matching_score": 1.2274911403656006,
325
+ "mean_text_rank": 1.5,
326
+ "median_text_rank": 1.0,
327
+ "r_precision": [
328
+ 0.71875,
329
+ 0.90625,
330
+ 0.9375
331
+ ],
332
+ "retrieval_batch_size": 32,
333
+ "retrieval_num_batches": 1,
334
+ "retrieval_num_samples": 32,
335
+ "retrieval_order_policy": "manifest_order",
336
+ "retrieval_unique_datasets_per_batch": {
337
+ "max": 1.0,
338
+ "mean": 1.0,
339
+ "min": 1.0,
340
+ "std": 0.0
341
+ }
342
+ },
343
+ "reference": {
344
+ "matching_distance": {
345
+ "max": 1.3188464641571045,
346
+ "mean": 1.2265774760135384,
347
+ "min": 1.1549649238586426,
348
+ "std": 0.04007899902186925
349
+ },
350
+ "matching_score": 1.2265775203704834,
351
+ "mean_text_rank": 1.84375,
352
+ "median_text_rank": 1.0,
353
+ "r_precision": [
354
+ 0.6875,
355
+ 0.84375,
356
+ 0.90625
357
+ ],
358
+ "retrieval_batch_size": 32,
359
+ "retrieval_num_batches": 1,
360
+ "retrieval_num_samples": 32,
361
+ "retrieval_order_policy": "manifest_order",
362
+ "retrieval_unique_datasets_per_batch": {
363
+ "max": 1.0,
364
+ "mean": 1.0,
365
+ "min": 1.0,
366
+ "std": 0.0
367
+ }
368
+ }
369
+ },
370
+ "transition": {
371
+ "auj": {
372
+ "max": 1.9905637383757522,
373
+ "mean": 1.7810514533823214,
374
+ "min": 1.5876606825310595,
375
+ "std": 0.06310559905267207
376
+ },
377
+ "body_auj": {
378
+ "max": 1.9905637383757522,
379
+ "mean": 1.7810514533823214,
380
+ "min": 1.5876606825310595,
381
+ "std": 0.06310559905267207
382
+ },
383
+ "body_pj": {
384
+ "max": 0.12257324159145355,
385
+ "mean": 0.04556534072210969,
386
+ "min": 0.0018733292818069458,
387
+ "std": 0.03456947191402774
388
+ },
389
+ "boundary_frames": [
390
+ 60
391
+ ],
392
+ "chunk_length": 60,
393
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
394
+ "num_frames": 120,
395
+ "num_samples": 43,
396
+ "pj": {
397
+ "max": 0.12257324159145355,
398
+ "mean": 0.04556534072210969,
399
+ "min": 0.0018733292818069458,
400
+ "std": 0.03456947191402774
401
+ },
402
+ "qpos_auj": {
403
+ "max": 20.530484854076903,
404
+ "mean": 6.420968679559986,
405
+ "min": 4.301883842105374,
406
+ "std": 4.107536056696693
407
+ },
408
+ "qpos_pj": {
409
+ "max": 6.2786277532577515,
410
+ "mean": 0.8628149387530907,
411
+ "min": 0.019462376832962036,
412
+ "std": 1.4660741357554568
413
+ },
414
+ "reference_body_jerk_level": 0.032168975994990016,
415
+ "reference_qpos_jerk_level": 0.07551584170589375,
416
+ "transition_window_length": 60,
417
+ "transition_windows": [
418
+ [
419
+ 30,
420
+ 90
421
+ ]
422
+ ]
423
+ }
424
+ },
425
+ "fitness_original_filtered_test": {
426
+ "embedding": {
427
+ "diversity_generated": 1.1245486499453385,
428
+ "motion_fid": 1.3721207787485414,
429
+ "motion_kid": {
430
+ "mean": 0.0019768632581005363,
431
+ "std": 0.0
432
+ },
433
+ "num_samples": 43
434
+ },
435
+ "num_samples": 43,
436
+ "physical": {
437
+ "body_jerk_mean": {
438
+ "max": 307.0351257324219,
439
+ "mean": 62.3401137174562,
440
+ "min": 5.94058895111084,
441
+ "std": 74.72415873290508
442
+ },
443
+ "contact_sliding_speed": {
444
+ "max": 1.9512882232666016,
445
+ "mean": 0.20101934357351342,
446
+ "min": 0.007332059554755688,
447
+ "std": 0.4436527359467069
448
+ },
449
+ "foot_ground_error": {
450
+ "max": 0.012699918821454048,
451
+ "mean": 0.006037447639459441,
452
+ "min": 0.0020615984685719013,
453
+ "std": 0.0021288537281194384
454
+ },
455
+ "num_samples": 43
456
+ },
457
+ "text": {
458
+ "generated": {
459
+ "matching_distance": {
460
+ "max": 1.460925579071045,
461
+ "mean": 1.3796154177466105,
462
+ "min": 1.268312692642212,
463
+ "std": 0.04377553385148069
464
+ },
465
+ "matching_score": 1.3796154260635376,
466
+ "mean_text_rank": 10.46875,
467
+ "median_text_rank": 11.0,
468
+ "r_precision": [
469
+ 0.125,
470
+ 0.125,
471
+ 0.125
472
+ ],
473
+ "retrieval_batch_size": 32,
474
+ "retrieval_num_batches": 1,
475
+ "retrieval_num_samples": 32,
476
+ "retrieval_order_policy": "manifest_order",
477
+ "retrieval_unique_datasets_per_batch": {
478
+ "max": 1.0,
479
+ "mean": 1.0,
480
+ "min": 1.0,
481
+ "std": 0.0
482
+ }
483
+ },
484
+ "reference": {
485
+ "matching_distance": {
486
+ "max": 1.3819284439086914,
487
+ "mean": 1.2803652868714444,
488
+ "min": 1.209172010421753,
489
+ "std": 0.03319129105747331
490
+ },
491
+ "matching_score": 1.280365228652954,
492
+ "mean_text_rank": 7.5,
493
+ "median_text_rank": 7.5,
494
+ "r_precision": [
495
+ 0.21875,
496
+ 0.25,
497
+ 0.3125
498
+ ],
499
+ "retrieval_batch_size": 32,
500
+ "retrieval_num_batches": 1,
501
+ "retrieval_num_samples": 32,
502
+ "retrieval_order_policy": "manifest_order",
503
+ "retrieval_unique_datasets_per_batch": {
504
+ "max": 1.0,
505
+ "mean": 1.0,
506
+ "min": 1.0,
507
+ "std": 0.0
508
+ }
509
+ }
510
+ },
511
+ "transition": {
512
+ "auj": {
513
+ "max": 2.385969114632759,
514
+ "mean": 1.8030387962130037,
515
+ "min": 1.4960776350777631,
516
+ "std": 0.11265973690452995
517
+ },
518
+ "body_auj": {
519
+ "max": 2.385969114632759,
520
+ "mean": 1.8030387962130037,
521
+ "min": 1.4960776350777631,
522
+ "std": 0.11265973690452995
523
+ },
524
+ "body_pj": {
525
+ "max": 0.21083378558978438,
526
+ "mean": 0.04393309076468265,
527
+ "min": 0.0029095709323883057,
528
+ "std": 0.0441904998385772
529
+ },
530
+ "boundary_frames": [
531
+ 60
532
+ ],
533
+ "chunk_length": 60,
534
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
535
+ "num_frames": 120,
536
+ "num_samples": 43,
537
+ "pj": {
538
+ "max": 0.21083378558978438,
539
+ "mean": 0.04393309076468265,
540
+ "min": 0.0029095709323883057,
541
+ "std": 0.0441904998385772
542
+ },
543
+ "qpos_auj": {
544
+ "max": 144.49321365365512,
545
+ "mean": 7.995620672684568,
546
+ "min": 4.299943787984878,
547
+ "std": 21.08138489740843
548
+ },
549
+ "qpos_pj": {
550
+ "max": 12.581249475479126,
551
+ "mean": 0.4969454543586539,
552
+ "min": 0.011551141738891602,
553
+ "std": 1.8728602685043632
554
+ },
555
+ "reference_body_jerk_level": 0.032168975994990016,
556
+ "reference_qpos_jerk_level": 0.07551584170589375,
557
+ "transition_window_length": 60,
558
+ "transition_windows": [
559
+ [
560
+ 30,
561
+ 90
562
+ ]
563
+ ]
564
+ }
565
+ },
566
+ "humanml_original_filtered_test": {
567
+ "embedding": {
568
+ "diversity_generated": 1.1726888731015561,
569
+ "motion_fid": 1.175771434491134,
570
+ "motion_kid": {
571
+ "mean": 0.0014073873525246583,
572
+ "std": 0.0
573
+ },
574
+ "num_samples": 43
575
+ },
576
+ "num_samples": 43,
577
+ "physical": {
578
+ "body_jerk_mean": {
579
+ "max": 246.43716430664062,
580
+ "mean": 68.01800821548285,
581
+ "min": 2.1355345249176025,
582
+ "std": 57.236756620842854
583
+ },
584
+ "contact_sliding_speed": {
585
+ "max": 1.9895073175430298,
586
+ "mean": 0.39400377310812473,
587
+ "min": 0.0,
588
+ "std": 0.46918049539700873
589
+ },
590
+ "foot_ground_error": {
591
+ "max": 1.477441668510437,
592
+ "mean": 0.04231151855594024,
593
+ "min": 0.002110379049554467,
594
+ "std": 0.2215764948289639
595
+ },
596
+ "num_samples": 43
597
+ },
598
+ "text": {
599
+ "generated": {
600
+ "matching_distance": {
601
+ "max": 1.430628776550293,
602
+ "mean": 1.3250223120977713,
603
+ "min": 1.2387126684188843,
604
+ "std": 0.04734495346885312
605
+ },
606
+ "matching_score": 1.3250223398208618,
607
+ "mean_text_rank": 6.71875,
608
+ "median_text_rank": 3.0,
609
+ "r_precision": [
610
+ 0.28125,
611
+ 0.46875,
612
+ 0.5625
613
+ ],
614
+ "retrieval_batch_size": 32,
615
+ "retrieval_num_batches": 1,
616
+ "retrieval_num_samples": 32,
617
+ "retrieval_order_policy": "manifest_order",
618
+ "retrieval_unique_datasets_per_batch": {
619
+ "max": 1.0,
620
+ "mean": 1.0,
621
+ "min": 1.0,
622
+ "std": 0.0
623
+ }
624
+ },
625
+ "reference": {
626
+ "matching_distance": {
627
+ "max": 1.3049654960632324,
628
+ "mean": 1.2212754654329876,
629
+ "min": 1.151880145072937,
630
+ "std": 0.033672752576929775
631
+ },
632
+ "matching_score": 1.2212754487991333,
633
+ "mean_text_rank": 2.0625,
634
+ "median_text_rank": 1.0,
635
+ "r_precision": [
636
+ 0.53125,
637
+ 0.78125,
638
+ 0.875
639
+ ],
640
+ "retrieval_batch_size": 32,
641
+ "retrieval_num_batches": 1,
642
+ "retrieval_num_samples": 32,
643
+ "retrieval_order_policy": "manifest_order",
644
+ "retrieval_unique_datasets_per_batch": {
645
+ "max": 1.0,
646
+ "mean": 1.0,
647
+ "min": 1.0,
648
+ "std": 0.0
649
+ }
650
+ }
651
+ },
652
+ "transition": {
653
+ "auj": {
654
+ "max": 2.2740369624934447,
655
+ "mean": 1.7931604724291397,
656
+ "min": 1.6542664254946329,
657
+ "std": 0.08686737366358381
658
+ },
659
+ "body_auj": {
660
+ "max": 2.2740369624934447,
661
+ "mean": 1.7931604724291397,
662
+ "min": 1.6542664254946329,
663
+ "std": 0.08686737366358381
664
+ },
665
+ "body_pj": {
666
+ "max": 0.21533825621008873,
667
+ "mean": 0.04874728411046225,
668
+ "min": 0.0004047006368637085,
669
+ "std": 0.04271094174830256
670
+ },
671
+ "boundary_frames": [
672
+ 60
673
+ ],
674
+ "chunk_length": 60,
675
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
676
+ "num_frames": 120,
677
+ "num_samples": 43,
678
+ "pj": {
679
+ "max": 0.21533825621008873,
680
+ "mean": 0.04874728411046225,
681
+ "min": 0.0004047006368637085,
682
+ "std": 0.04271094174830256
683
+ },
684
+ "qpos_auj": {
685
+ "max": 10.246992179070007,
686
+ "mean": 5.0103711591434354,
687
+ "min": 4.300404030837447,
688
+ "std": 1.3281499654605524
689
+ },
690
+ "qpos_pj": {
691
+ "max": 2.85389244556427,
692
+ "mean": 0.3255547105367187,
693
+ "min": 0.0028911828994750977,
694
+ "std": 0.5239349095283596
695
+ },
696
+ "reference_body_jerk_level": 0.032168975994990016,
697
+ "reference_qpos_jerk_level": 0.07551584170589375,
698
+ "transition_window_length": 60,
699
+ "transition_windows": [
700
+ [
701
+ 30,
702
+ 90
703
+ ]
704
+ ]
705
+ }
706
+ },
707
+ "idea400_original_filtered_test": {
708
+ "embedding": {
709
+ "diversity_generated": 1.1694871863472327,
710
+ "motion_fid": 1.424413235305793,
711
+ "motion_kid": {
712
+ "mean": 0.003020059186301882,
713
+ "std": 0.0
714
+ },
715
+ "num_samples": 43
716
+ },
717
+ "num_samples": 43,
718
+ "physical": {
719
+ "body_jerk_mean": {
720
+ "max": 185.40188598632812,
721
+ "mean": 48.86681900467983,
722
+ "min": 2.1356678009033203,
723
+ "std": 52.217468326814576
724
+ },
725
+ "contact_sliding_speed": {
726
+ "max": 1.2346510887145996,
727
+ "mean": 0.27460066539365363,
728
+ "min": 0.0036743313539773226,
729
+ "std": 0.3540692147671962
730
+ },
731
+ "foot_ground_error": {
732
+ "max": 0.01912386529147625,
733
+ "mean": 0.0067057522327834095,
734
+ "min": 0.00249672238714993,
735
+ "std": 0.003613244239824124
736
+ },
737
+ "num_samples": 43
738
+ },
739
+ "text": {
740
+ "generated": {
741
+ "matching_distance": {
742
+ "max": 1.4311089515686035,
743
+ "mean": 1.3534246184105096,
744
+ "min": 1.2549099922180176,
745
+ "std": 0.040087949446493104
746
+ },
747
+ "matching_score": 1.3534245491027832,
748
+ "mean_text_rank": 8.5,
749
+ "median_text_rank": 5.0,
750
+ "r_precision": [
751
+ 0.1875,
752
+ 0.21875,
753
+ 0.3125
754
+ ],
755
+ "retrieval_batch_size": 32,
756
+ "retrieval_num_batches": 1,
757
+ "retrieval_num_samples": 32,
758
+ "retrieval_order_policy": "manifest_order",
759
+ "retrieval_unique_datasets_per_batch": {
760
+ "max": 1.0,
761
+ "mean": 1.0,
762
+ "min": 1.0,
763
+ "std": 0.0
764
+ }
765
+ },
766
+ "reference": {
767
+ "matching_distance": {
768
+ "max": 1.3933742046356201,
769
+ "mean": 1.253553493078365,
770
+ "min": 1.1590988636016846,
771
+ "std": 0.0395202842188391
772
+ },
773
+ "matching_score": 1.2535533905029297,
774
+ "mean_text_rank": 5.5625,
775
+ "median_text_rank": 4.0,
776
+ "r_precision": [
777
+ 0.21875,
778
+ 0.4375,
779
+ 0.46875
780
+ ],
781
+ "retrieval_batch_size": 32,
782
+ "retrieval_num_batches": 1,
783
+ "retrieval_num_samples": 32,
784
+ "retrieval_order_policy": "manifest_order",
785
+ "retrieval_unique_datasets_per_batch": {
786
+ "max": 1.0,
787
+ "mean": 1.0,
788
+ "min": 1.0,
789
+ "std": 0.0
790
+ }
791
+ }
792
+ },
793
+ "transition": {
794
+ "auj": {
795
+ "max": 1.9275732800142191,
796
+ "mean": 1.7972521730343467,
797
+ "min": 1.582865984803458,
798
+ "std": 0.06215447215088484
799
+ },
800
+ "body_auj": {
801
+ "max": 1.9275732800142191,
802
+ "mean": 1.7972521730343467,
803
+ "min": 1.582865984803458,
804
+ "std": 0.06215447215088484
805
+ },
806
+ "body_pj": {
807
+ "max": 0.19542989134788513,
808
+ "mean": 0.037036049071439474,
809
+ "min": 0.00043263286352157593,
810
+ "std": 0.04102339681952739
811
+ },
812
+ "boundary_frames": [
813
+ 60
814
+ ],
815
+ "chunk_length": 60,
816
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
817
+ "num_frames": 120,
818
+ "num_samples": 43,
819
+ "pj": {
820
+ "max": 0.19542989134788513,
821
+ "mean": 0.037036049071439474,
822
+ "min": 0.00043263286352157593,
823
+ "std": 0.04102339681952739
824
+ },
825
+ "qpos_auj": {
826
+ "max": 35.02763120426745,
827
+ "mean": 5.4219513973630304,
828
+ "min": 4.3006585100743475,
829
+ "std": 4.6576733588000225
830
+ },
831
+ "qpos_pj": {
832
+ "max": 12.468976974487305,
833
+ "mean": 0.5249752198849523,
834
+ "min": 0.002636820077896118,
835
+ "std": 1.8992311316861998
836
+ },
837
+ "reference_body_jerk_level": 0.032168975994990016,
838
+ "reference_qpos_jerk_level": 0.07551584170589375,
839
+ "transition_window_length": 60,
840
+ "transition_windows": [
841
+ [
842
+ 30,
843
+ 90
844
+ ]
845
+ ]
846
+ }
847
+ },
848
+ "lafan1_original_filtered_test": {
849
+ "embedding": {
850
+ "diversity_generated": 1.065131755766148,
851
+ "motion_fid": 0.9163817643240386,
852
+ "motion_kid": {
853
+ "mean": 0.001431817572137728,
854
+ "std": 0.0
855
+ },
856
+ "num_samples": 43
857
+ },
858
+ "num_samples": 43,
859
+ "physical": {
860
+ "body_jerk_mean": {
861
+ "max": 531.6823120117188,
862
+ "mean": 164.51493027043898,
863
+ "min": 6.81342077255249,
864
+ "std": 135.10997614115757
865
+ },
866
+ "contact_sliding_speed": {
867
+ "max": 2.914012908935547,
868
+ "mean": 0.8355859869354686,
869
+ "min": 0.008567381650209427,
870
+ "std": 0.7517330338413282
871
+ },
872
+ "foot_ground_error": {
873
+ "max": 0.013226441107690334,
874
+ "mean": 0.0067459214966051105,
875
+ "min": 0.0020911949686706066,
876
+ "std": 0.002395367031927562
877
+ },
878
+ "num_samples": 43
879
+ },
880
+ "text": {
881
+ "generated": {
882
+ "matching_distance": {
883
+ "max": 1.3756978511810303,
884
+ "mean": 1.2848902968473213,
885
+ "min": 1.22147798538208,
886
+ "std": 0.03506700337563933
887
+ },
888
+ "matching_score": 1.2848902940750122,
889
+ "mean_text_rank": 6.78125,
890
+ "median_text_rank": 5.0,
891
+ "r_precision": [
892
+ 0.375,
893
+ 0.40625,
894
+ 0.46875
895
+ ],
896
+ "retrieval_batch_size": 32,
897
+ "retrieval_num_batches": 1,
898
+ "retrieval_num_samples": 32,
899
+ "retrieval_order_policy": "manifest_order",
900
+ "retrieval_unique_datasets_per_batch": {
901
+ "max": 1.0,
902
+ "mean": 1.0,
903
+ "min": 1.0,
904
+ "std": 0.0
905
+ }
906
+ },
907
+ "reference": {
908
+ "matching_distance": {
909
+ "max": 1.3601514101028442,
910
+ "mean": 1.262607499610546,
911
+ "min": 1.1779136657714844,
912
+ "std": 0.036237216916877986
913
+ },
914
+ "matching_score": 1.262607455253601,
915
+ "mean_text_rank": 7.34375,
916
+ "median_text_rank": 4.0,
917
+ "r_precision": [
918
+ 0.28125,
919
+ 0.34375,
920
+ 0.46875
921
+ ],
922
+ "retrieval_batch_size": 32,
923
+ "retrieval_num_batches": 1,
924
+ "retrieval_num_samples": 32,
925
+ "retrieval_order_policy": "manifest_order",
926
+ "retrieval_unique_datasets_per_batch": {
927
+ "max": 1.0,
928
+ "mean": 1.0,
929
+ "min": 1.0,
930
+ "std": 0.0
931
+ }
932
+ }
933
+ },
934
+ "transition": {
935
+ "auj": {
936
+ "max": 4.683557439704222,
937
+ "mean": 1.9927798923981312,
938
+ "min": 1.7141404278200993,
939
+ "std": 0.5034205327935629
940
+ },
941
+ "body_auj": {
942
+ "max": 4.683557439704222,
943
+ "mean": 1.9927798923981312,
944
+ "min": 1.7141404278200993,
945
+ "std": 0.5034205327935629
946
+ },
947
+ "body_pj": {
948
+ "max": 0.8904187381267548,
949
+ "mean": 0.11073192904248487,
950
+ "min": 0.0009953081607818604,
951
+ "std": 0.13824281051787887
952
+ },
953
+ "boundary_frames": [
954
+ 60
955
+ ],
956
+ "chunk_length": 60,
957
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
958
+ "num_frames": 120,
959
+ "num_samples": 43,
960
+ "pj": {
961
+ "max": 0.8904187381267548,
962
+ "mean": 0.11073192904248487,
963
+ "min": 0.0009953081607818604,
964
+ "std": 0.13824281051787887
965
+ },
966
+ "qpos_auj": {
967
+ "max": 43.07620444184711,
968
+ "mean": 8.566258157005608,
969
+ "min": 4.297486459604036,
970
+ "std": 7.636360765758662
971
+ },
972
+ "qpos_pj": {
973
+ "max": 12.626699447631836,
974
+ "mean": 1.2561438795352398,
975
+ "min": 0.00521540641784668,
976
+ "std": 2.344095482079285
977
+ },
978
+ "reference_body_jerk_level": 0.032168975994990016,
979
+ "reference_qpos_jerk_level": 0.07551584170589375,
980
+ "transition_window_length": 60,
981
+ "transition_windows": [
982
+ [
983
+ 30,
984
+ 90
985
+ ]
986
+ ]
987
+ }
988
+ },
989
+ "motiongv_original_filtered_test": {
990
+ "embedding": {
991
+ "diversity_generated": 1.151175892575208,
992
+ "motion_fid": 1.4664699422882885,
993
+ "motion_kid": {
994
+ "mean": 0.0018693371237699097,
995
+ "std": 0.0
996
+ },
997
+ "num_samples": 43
998
+ },
999
+ "num_samples": 43,
1000
+ "physical": {
1001
+ "body_jerk_mean": {
1002
+ "max": 391.6789245605469,
1003
+ "mean": 82.85689859612043,
1004
+ "min": 3.0611748695373535,
1005
+ "std": 95.17589490486428
1006
+ },
1007
+ "contact_sliding_speed": {
1008
+ "max": 1.5828344821929932,
1009
+ "mean": 0.17748946919604097,
1010
+ "min": 0.006206035614013672,
1011
+ "std": 0.2848895833067525
1012
+ },
1013
+ "foot_ground_error": {
1014
+ "max": 0.05884654074907303,
1015
+ "mean": 0.00999552955266175,
1016
+ "min": 0.0020414358004927635,
1017
+ "std": 0.009597411811076375
1018
+ },
1019
+ "num_samples": 43
1020
+ },
1021
+ "text": {
1022
+ "generated": {
1023
+ "matching_distance": {
1024
+ "max": 1.472267508506775,
1025
+ "mean": 1.3855077416397805,
1026
+ "min": 1.2616697549819946,
1027
+ "std": 0.04268846702374364
1028
+ },
1029
+ "matching_score": 1.3855078220367432,
1030
+ "mean_text_rank": 6.71875,
1031
+ "median_text_rank": 3.5,
1032
+ "r_precision": [
1033
+ 0.28125,
1034
+ 0.40625,
1035
+ 0.5
1036
+ ],
1037
+ "retrieval_batch_size": 32,
1038
+ "retrieval_num_batches": 1,
1039
+ "retrieval_num_samples": 32,
1040
+ "retrieval_order_policy": "manifest_order",
1041
+ "retrieval_unique_datasets_per_batch": {
1042
+ "max": 1.0,
1043
+ "mean": 1.0,
1044
+ "min": 1.0,
1045
+ "std": 0.0
1046
+ }
1047
+ },
1048
+ "reference": {
1049
+ "matching_distance": {
1050
+ "max": 1.355556845664978,
1051
+ "mean": 1.2528588660927706,
1052
+ "min": 1.1668779850006104,
1053
+ "std": 0.04499687839541838
1054
+ },
1055
+ "matching_score": 1.2528588771820068,
1056
+ "mean_text_rank": 1.9375,
1057
+ "median_text_rank": 1.0,
1058
+ "r_precision": [
1059
+ 0.65625,
1060
+ 0.8125,
1061
+ 0.8125
1062
+ ],
1063
+ "retrieval_batch_size": 32,
1064
+ "retrieval_num_batches": 1,
1065
+ "retrieval_num_samples": 32,
1066
+ "retrieval_order_policy": "manifest_order",
1067
+ "retrieval_unique_datasets_per_batch": {
1068
+ "max": 1.0,
1069
+ "mean": 1.0,
1070
+ "min": 1.0,
1071
+ "std": 0.0
1072
+ }
1073
+ }
1074
+ },
1075
+ "transition": {
1076
+ "auj": {
1077
+ "max": 3.7732563196112654,
1078
+ "mean": 1.877907262608954,
1079
+ "min": 1.4860971376024386,
1080
+ "std": 0.3515215707606796
1081
+ },
1082
+ "body_auj": {
1083
+ "max": 3.7732563196112654,
1084
+ "mean": 1.877907262608954,
1085
+ "min": 1.4860971376024386,
1086
+ "std": 0.3515215707606796
1087
+ },
1088
+ "body_pj": {
1089
+ "max": 0.5607699304819107,
1090
+ "mean": 0.059313230828360416,
1091
+ "min": 0.0006179101765155792,
1092
+ "std": 0.09926593614052223
1093
+ },
1094
+ "boundary_frames": [
1095
+ 60
1096
+ ],
1097
+ "chunk_length": 60,
1098
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1099
+ "num_frames": 120,
1100
+ "num_samples": 43,
1101
+ "pj": {
1102
+ "max": 0.5607699304819107,
1103
+ "mean": 0.059313230828360416,
1104
+ "min": 0.0006179101765155792,
1105
+ "std": 0.09926593614052223
1106
+ },
1107
+ "qpos_auj": {
1108
+ "max": 62.14833038614482,
1109
+ "mean": 7.156893734841608,
1110
+ "min": 4.296862431809634,
1111
+ "std": 10.065915962574447
1112
+ },
1113
+ "qpos_pj": {
1114
+ "max": 12.888956546783447,
1115
+ "mean": 0.8218644026580245,
1116
+ "min": 0.003064531832933426,
1117
+ "std": 2.6184541571050173
1118
+ },
1119
+ "reference_body_jerk_level": 0.032168975994990016,
1120
+ "reference_qpos_jerk_level": 0.07551584170589375,
1121
+ "transition_window_length": 60,
1122
+ "transition_windows": [
1123
+ [
1124
+ 30,
1125
+ 90
1126
+ ]
1127
+ ]
1128
+ }
1129
+ },
1130
+ "motionllama_original_filtered_test": {
1131
+ "embedding": {
1132
+ "diversity_generated": 1.1930282606242615,
1133
+ "motion_fid": 1.3093164078052215,
1134
+ "motion_kid": {
1135
+ "mean": 0.001977269528855352,
1136
+ "std": 0.0
1137
+ },
1138
+ "num_samples": 42
1139
+ },
1140
+ "num_samples": 42,
1141
+ "physical": {
1142
+ "body_jerk_mean": {
1143
+ "max": 185.41676330566406,
1144
+ "mean": 17.548739166486833,
1145
+ "min": 2.4985158443450928,
1146
+ "std": 29.38218129714949
1147
+ },
1148
+ "contact_sliding_speed": {
1149
+ "max": 1.210018515586853,
1150
+ "mean": 0.05310293933421019,
1151
+ "min": 0.004471967928111553,
1152
+ "std": 0.18190039828460755
1153
+ },
1154
+ "foot_ground_error": {
1155
+ "max": 0.11203083395957947,
1156
+ "mean": 0.008613292068254114,
1157
+ "min": 0.0008450289606116712,
1158
+ "std": 0.016888853227158128
1159
+ },
1160
+ "num_samples": 42
1161
+ },
1162
+ "text": {
1163
+ "generated": {
1164
+ "matching_distance": {
1165
+ "max": 1.4584448337554932,
1166
+ "mean": 1.3611151774724324,
1167
+ "min": 1.2479358911514282,
1168
+ "std": 0.0521328721850203
1169
+ },
1170
+ "matching_score": 1.3611150979995728,
1171
+ "mean_text_rank": 7.75,
1172
+ "median_text_rank": 4.0,
1173
+ "r_precision": [
1174
+ 0.1875,
1175
+ 0.34375,
1176
+ 0.46875
1177
+ ],
1178
+ "retrieval_batch_size": 32,
1179
+ "retrieval_num_batches": 1,
1180
+ "retrieval_num_samples": 32,
1181
+ "retrieval_order_policy": "manifest_order",
1182
+ "retrieval_unique_datasets_per_batch": {
1183
+ "max": 1.0,
1184
+ "mean": 1.0,
1185
+ "min": 1.0,
1186
+ "std": 0.0
1187
+ }
1188
+ },
1189
+ "reference": {
1190
+ "matching_distance": {
1191
+ "max": 1.344225287437439,
1192
+ "mean": 1.2109532526561193,
1193
+ "min": 1.1528350114822388,
1194
+ "std": 0.03675615838449094
1195
+ },
1196
+ "matching_score": 1.2109531164169312,
1197
+ "mean_text_rank": 2.40625,
1198
+ "median_text_rank": 1.0,
1199
+ "r_precision": [
1200
+ 0.53125,
1201
+ 0.6875,
1202
+ 0.8125
1203
+ ],
1204
+ "retrieval_batch_size": 32,
1205
+ "retrieval_num_batches": 1,
1206
+ "retrieval_num_samples": 32,
1207
+ "retrieval_order_policy": "manifest_order",
1208
+ "retrieval_unique_datasets_per_batch": {
1209
+ "max": 1.0,
1210
+ "mean": 1.0,
1211
+ "min": 1.0,
1212
+ "std": 0.0
1213
+ }
1214
+ }
1215
+ },
1216
+ "transition": {
1217
+ "auj": {
1218
+ "max": 1.8308297701004421,
1219
+ "mean": 1.8203518604107993,
1220
+ "min": 1.7468128703884296,
1221
+ "std": 0.015218604769691082
1222
+ },
1223
+ "body_auj": {
1224
+ "max": 1.8308297701004421,
1225
+ "mean": 1.8203518604107993,
1226
+ "min": 1.7468128703884296,
1227
+ "std": 0.015218604769691082
1228
+ },
1229
+ "body_pj": {
1230
+ "max": 0.09689441323280334,
1231
+ "mean": 0.011014723008321155,
1232
+ "min": 0.0005841478705406189,
1233
+ "std": 0.017570536124353737
1234
+ },
1235
+ "boundary_frames": [
1236
+ 60
1237
+ ],
1238
+ "chunk_length": 60,
1239
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1240
+ "num_frames": 120,
1241
+ "num_samples": 42,
1242
+ "pj": {
1243
+ "max": 0.09689441323280334,
1244
+ "mean": 0.011014723008321155,
1245
+ "min": 0.0005841478705406189,
1246
+ "std": 0.017570536124353737
1247
+ },
1248
+ "qpos_auj": {
1249
+ "max": 9.52280969330879,
1250
+ "mean": 4.517935173208866,
1251
+ "min": 4.3012650981623235,
1252
+ "std": 0.8728498919105634
1253
+ },
1254
+ "qpos_pj": {
1255
+ "max": 2.8352200984954834,
1256
+ "mean": 0.1325104517579478,
1257
+ "min": 0.0033388733863830566,
1258
+ "std": 0.4389731704134287
1259
+ },
1260
+ "reference_body_jerk_level": 0.032168975994990016,
1261
+ "reference_qpos_jerk_level": 0.07551584170589375,
1262
+ "transition_window_length": 60,
1263
+ "transition_windows": [
1264
+ [
1265
+ 30,
1266
+ 90
1267
+ ]
1268
+ ]
1269
+ }
1270
+ },
1271
+ "omomo_original_filtered_test": {
1272
+ "embedding": {
1273
+ "diversity_generated": 1.1345490872611939,
1274
+ "motion_fid": 1.2325407801917518,
1275
+ "motion_kid": {
1276
+ "mean": 0.002298161219308703,
1277
+ "std": 0.0
1278
+ },
1279
+ "num_samples": 42
1280
+ },
1281
+ "num_samples": 42,
1282
+ "physical": {
1283
+ "body_jerk_mean": {
1284
+ "max": 215.9715118408203,
1285
+ "mean": 64.15830364681426,
1286
+ "min": 10.436027526855469,
1287
+ "std": 44.76208781970577
1288
+ },
1289
+ "contact_sliding_speed": {
1290
+ "max": 1.5735926628112793,
1291
+ "mean": 0.47663801983885823,
1292
+ "min": 0.015720246359705925,
1293
+ "std": 0.4178593234495149
1294
+ },
1295
+ "foot_ground_error": {
1296
+ "max": 0.01301527488976717,
1297
+ "mean": 0.006088023423217237,
1298
+ "min": 0.0020812139846384525,
1299
+ "std": 0.002377843957042252
1300
+ },
1301
+ "num_samples": 42
1302
+ },
1303
+ "text": {
1304
+ "generated": {
1305
+ "matching_distance": {
1306
+ "max": 1.433463454246521,
1307
+ "mean": 1.305754039968763,
1308
+ "min": 1.2346299886703491,
1309
+ "std": 0.042840413554061436
1310
+ },
1311
+ "matching_score": 1.3057539463043213,
1312
+ "mean_text_rank": 8.6875,
1313
+ "median_text_rank": 5.5,
1314
+ "r_precision": [
1315
+ 0.25,
1316
+ 0.3125,
1317
+ 0.40625
1318
+ ],
1319
+ "retrieval_batch_size": 32,
1320
+ "retrieval_num_batches": 1,
1321
+ "retrieval_num_samples": 32,
1322
+ "retrieval_order_policy": "manifest_order",
1323
+ "retrieval_unique_datasets_per_batch": {
1324
+ "max": 1.0,
1325
+ "mean": 1.0,
1326
+ "min": 1.0,
1327
+ "std": 0.0
1328
+ }
1329
+ },
1330
+ "reference": {
1331
+ "matching_distance": {
1332
+ "max": 1.3671505451202393,
1333
+ "mean": 1.269894642489297,
1334
+ "min": 1.2092769145965576,
1335
+ "std": 0.03030912014653846
1336
+ },
1337
+ "matching_score": 1.2698945999145508,
1338
+ "mean_text_rank": 8.40625,
1339
+ "median_text_rank": 6.5,
1340
+ "r_precision": [
1341
+ 0.15625,
1342
+ 0.21875,
1343
+ 0.25
1344
+ ],
1345
+ "retrieval_batch_size": 32,
1346
+ "retrieval_num_batches": 1,
1347
+ "retrieval_num_samples": 32,
1348
+ "retrieval_order_policy": "manifest_order",
1349
+ "retrieval_unique_datasets_per_batch": {
1350
+ "max": 1.0,
1351
+ "mean": 1.0,
1352
+ "min": 1.0,
1353
+ "std": 0.0
1354
+ }
1355
+ }
1356
+ },
1357
+ "transition": {
1358
+ "auj": {
1359
+ "max": 1.8620231315843276,
1360
+ "mean": 1.7725733755803437,
1361
+ "min": 1.6676089530947735,
1362
+ "std": 0.049378376046624706
1363
+ },
1364
+ "body_auj": {
1365
+ "max": 1.8620231315843276,
1366
+ "mean": 1.7725733755803437,
1367
+ "min": 1.6676089530947735,
1368
+ "std": 0.049378376046624706
1369
+ },
1370
+ "body_pj": {
1371
+ "max": 0.1809138059616089,
1372
+ "mean": 0.04543182488885664,
1373
+ "min": 0.004288120195269585,
1374
+ "std": 0.03565515372849116
1375
+ },
1376
+ "boundary_frames": [
1377
+ 60
1378
+ ],
1379
+ "chunk_length": 60,
1380
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1381
+ "num_frames": 120,
1382
+ "num_samples": 42,
1383
+ "pj": {
1384
+ "max": 0.1809138059616089,
1385
+ "mean": 0.04543182488885664,
1386
+ "min": 0.004288120195269585,
1387
+ "std": 0.03565515372849116
1388
+ },
1389
+ "qpos_auj": {
1390
+ "max": 28.471612076712674,
1391
+ "mean": 5.603556235023785,
1392
+ "min": 4.300355892869041,
1393
+ "std": 3.8283692334934805
1394
+ },
1395
+ "qpos_pj": {
1396
+ "max": 11.778656005859375,
1397
+ "mean": 0.6776692000927315,
1398
+ "min": 0.007280692458152771,
1399
+ "std": 1.8678563037401907
1400
+ },
1401
+ "reference_body_jerk_level": 0.032168975994990016,
1402
+ "reference_qpos_jerk_level": 0.07551584170589375,
1403
+ "transition_window_length": 60,
1404
+ "transition_windows": [
1405
+ [
1406
+ 30,
1407
+ 90
1408
+ ]
1409
+ ]
1410
+ }
1411
+ },
1412
+ "permo_original_filtered_test": {
1413
+ "embedding": {
1414
+ "diversity_generated": 1.0877195287239692,
1415
+ "motion_fid": 1.1566815401268762,
1416
+ "motion_kid": {
1417
+ "mean": 0.0022956662767517066,
1418
+ "std": 0.0
1419
+ },
1420
+ "num_samples": 42
1421
+ },
1422
+ "num_samples": 42,
1423
+ "physical": {
1424
+ "body_jerk_mean": {
1425
+ "max": 263.7925720214844,
1426
+ "mean": 76.4998429048629,
1427
+ "min": 4.8154168128967285,
1428
+ "std": 69.10691261564781
1429
+ },
1430
+ "contact_sliding_speed": {
1431
+ "max": 1.7778644561767578,
1432
+ "mean": 0.31765077484860305,
1433
+ "min": 0.009715426713228226,
1434
+ "std": 0.403836893708808
1435
+ },
1436
+ "foot_ground_error": {
1437
+ "max": 0.025011582300066948,
1438
+ "mean": 0.007566677107076559,
1439
+ "min": 0.0026211035437881947,
1440
+ "std": 0.004112016751390115
1441
+ },
1442
+ "num_samples": 42
1443
+ },
1444
+ "text": {
1445
+ "generated": {
1446
+ "matching_distance": {
1447
+ "max": 1.3902548551559448,
1448
+ "mean": 1.3313227948688326,
1449
+ "min": 1.2616204023361206,
1450
+ "std": 0.0340989370710566
1451
+ },
1452
+ "matching_score": 1.3313229084014893,
1453
+ "mean_text_rank": 4.84375,
1454
+ "median_text_rank": 4.0,
1455
+ "r_precision": [
1456
+ 0.375,
1457
+ 0.40625,
1458
+ 0.4375
1459
+ ],
1460
+ "retrieval_batch_size": 32,
1461
+ "retrieval_num_batches": 1,
1462
+ "retrieval_num_samples": 32,
1463
+ "retrieval_order_policy": "manifest_order",
1464
+ "retrieval_unique_datasets_per_batch": {
1465
+ "max": 1.0,
1466
+ "mean": 1.0,
1467
+ "min": 1.0,
1468
+ "std": 0.0
1469
+ }
1470
+ },
1471
+ "reference": {
1472
+ "matching_distance": {
1473
+ "max": 1.3314403295516968,
1474
+ "mean": 1.2064374969119118,
1475
+ "min": 1.1483595371246338,
1476
+ "std": 0.03169944929872626
1477
+ },
1478
+ "matching_score": 1.2064374685287476,
1479
+ "mean_text_rank": 4.34375,
1480
+ "median_text_rank": 3.0,
1481
+ "r_precision": [
1482
+ 0.25,
1483
+ 0.4375,
1484
+ 0.53125
1485
+ ],
1486
+ "retrieval_batch_size": 32,
1487
+ "retrieval_num_batches": 1,
1488
+ "retrieval_num_samples": 32,
1489
+ "retrieval_order_policy": "manifest_order",
1490
+ "retrieval_unique_datasets_per_batch": {
1491
+ "max": 1.0,
1492
+ "mean": 1.0,
1493
+ "min": 1.0,
1494
+ "std": 0.0
1495
+ }
1496
+ }
1497
+ },
1498
+ "transition": {
1499
+ "auj": {
1500
+ "max": 2.781857592675066,
1501
+ "mean": 1.8609303950500373,
1502
+ "min": 1.5932010662360152,
1503
+ "std": 0.2132839801381445
1504
+ },
1505
+ "body_auj": {
1506
+ "max": 2.781857592675066,
1507
+ "mean": 1.8609303950500373,
1508
+ "min": 1.5932010662360152,
1509
+ "std": 0.2132839801381445
1510
+ },
1511
+ "body_pj": {
1512
+ "max": 0.2385309636592865,
1513
+ "mean": 0.0728943720903425,
1514
+ "min": 0.0014599980786442757,
1515
+ "std": 0.0661067180136749
1516
+ },
1517
+ "boundary_frames": [
1518
+ 60
1519
+ ],
1520
+ "chunk_length": 60,
1521
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1522
+ "num_frames": 120,
1523
+ "num_samples": 42,
1524
+ "pj": {
1525
+ "max": 0.2385309636592865,
1526
+ "mean": 0.0728943720903425,
1527
+ "min": 0.0014599980786442757,
1528
+ "std": 0.0661067180136749
1529
+ },
1530
+ "qpos_auj": {
1531
+ "max": 12.99549435606882,
1532
+ "mean": 5.456390068404779,
1533
+ "min": 4.300785458770909,
1534
+ "std": 1.9640080094470993
1535
+ },
1536
+ "qpos_pj": {
1537
+ "max": 3.9514918252825737,
1538
+ "mean": 0.4679690366056526,
1539
+ "min": 0.006184376776218414,
1540
+ "std": 0.7646300858697805
1541
+ },
1542
+ "reference_body_jerk_level": 0.032168975994990016,
1543
+ "reference_qpos_jerk_level": 0.07551584170589375,
1544
+ "transition_window_length": 60,
1545
+ "transition_windows": [
1546
+ [
1547
+ 30,
1548
+ 90
1549
+ ]
1550
+ ]
1551
+ }
1552
+ },
1553
+ "snapmogen_original_filtered_test": {
1554
+ "embedding": {
1555
+ "diversity_generated": 1.2069637839684089,
1556
+ "motion_fid": 1.3497101089407968,
1557
+ "motion_kid": {
1558
+ "mean": 0.0020190263621038262,
1559
+ "std": 0.0
1560
+ },
1561
+ "num_samples": 42
1562
+ },
1563
+ "num_samples": 42,
1564
+ "physical": {
1565
+ "body_jerk_mean": {
1566
+ "max": 228.76815795898438,
1567
+ "mean": 39.739420181229,
1568
+ "min": 2.503999710083008,
1569
+ "std": 55.302429574414276
1570
+ },
1571
+ "contact_sliding_speed": {
1572
+ "max": 1.628731369972229,
1573
+ "mean": 0.2116111345655684,
1574
+ "min": 0.006765882950276136,
1575
+ "std": 0.36438021456289116
1576
+ },
1577
+ "foot_ground_error": {
1578
+ "max": 0.011111666448414326,
1579
+ "mean": 0.005569282187415021,
1580
+ "min": 0.0005915897199884057,
1581
+ "std": 0.0024525098901194337
1582
+ },
1583
+ "num_samples": 42
1584
+ },
1585
+ "text": {
1586
+ "generated": {
1587
+ "matching_distance": {
1588
+ "max": 1.4769279956817627,
1589
+ "mean": 1.377476453781128,
1590
+ "min": 1.2130833864212036,
1591
+ "std": 0.046379623782035695
1592
+ },
1593
+ "matching_score": 1.377476453781128,
1594
+ "mean_text_rank": 6.84375,
1595
+ "median_text_rank": 4.0,
1596
+ "r_precision": [
1597
+ 0.34375,
1598
+ 0.4375,
1599
+ 0.46875
1600
+ ],
1601
+ "retrieval_batch_size": 32,
1602
+ "retrieval_num_batches": 1,
1603
+ "retrieval_num_samples": 32,
1604
+ "retrieval_order_policy": "manifest_order",
1605
+ "retrieval_unique_datasets_per_batch": {
1606
+ "max": 1.0,
1607
+ "mean": 1.0,
1608
+ "min": 1.0,
1609
+ "std": 0.0
1610
+ }
1611
+ },
1612
+ "reference": {
1613
+ "matching_distance": {
1614
+ "max": 1.2890117168426514,
1615
+ "mean": 1.2102941047577631,
1616
+ "min": 1.1285470724105835,
1617
+ "std": 0.04065653924035539
1618
+ },
1619
+ "matching_score": 1.2102941274642944,
1620
+ "mean_text_rank": 1.75,
1621
+ "median_text_rank": 1.0,
1622
+ "r_precision": [
1623
+ 0.75,
1624
+ 0.8125,
1625
+ 0.84375
1626
+ ],
1627
+ "retrieval_batch_size": 32,
1628
+ "retrieval_num_batches": 1,
1629
+ "retrieval_num_samples": 32,
1630
+ "retrieval_order_policy": "manifest_order",
1631
+ "retrieval_unique_datasets_per_batch": {
1632
+ "max": 1.0,
1633
+ "mean": 1.0,
1634
+ "min": 1.0,
1635
+ "std": 0.0
1636
+ }
1637
+ }
1638
+ },
1639
+ "transition": {
1640
+ "auj": {
1641
+ "max": 5.563047105616452,
1642
+ "mean": 1.9004562439177437,
1643
+ "min": 1.720805086064156,
1644
+ "std": 0.5739356354454959
1645
+ },
1646
+ "body_auj": {
1647
+ "max": 5.563047105616452,
1648
+ "mean": 1.9004562439177437,
1649
+ "min": 1.720805086064156,
1650
+ "std": 0.5739356354454959
1651
+ },
1652
+ "body_pj": {
1653
+ "max": 1.3313361704349518,
1654
+ "mean": 0.058183968610440694,
1655
+ "min": 0.0004454776644706726,
1656
+ "std": 0.20293248270696373
1657
+ },
1658
+ "boundary_frames": [
1659
+ 60
1660
+ ],
1661
+ "chunk_length": 60,
1662
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1663
+ "num_frames": 120,
1664
+ "num_samples": 42,
1665
+ "pj": {
1666
+ "max": 1.3313361704349518,
1667
+ "mean": 0.058183968610440694,
1668
+ "min": 0.0004454776644706726,
1669
+ "std": 0.20293248270696373
1670
+ },
1671
+ "qpos_auj": {
1672
+ "max": 52.49802010814777,
1673
+ "mean": 6.607324879388726,
1674
+ "min": 4.299905554166392,
1675
+ "std": 8.780425573903713
1676
+ },
1677
+ "qpos_pj": {
1678
+ "max": 12.550551891326904,
1679
+ "mean": 0.81523050702942,
1680
+ "min": 0.0026228129863739014,
1681
+ "std": 2.5778904370957694
1682
+ },
1683
+ "reference_body_jerk_level": 0.032168975994990016,
1684
+ "reference_qpos_jerk_level": 0.07551584170589375,
1685
+ "transition_window_length": 60,
1686
+ "transition_windows": [
1687
+ [
1688
+ 30,
1689
+ 90
1690
+ ]
1691
+ ]
1692
+ }
1693
+ }
1694
+ }
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/embedding_metrics.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "diversity_generated": 1.185870855502696,
3
+ "embedding_frames": 120,
4
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_040000.pt",
5
+ "matching_score_generated": 1.3311593532562256,
6
+ "matching_score_reference": 1.244696855545044,
7
+ "motion_fid": 0.5886926257750928,
8
+ "motion_key": "body_pos_local",
9
+ "motion_kid": {
10
+ "mean": 0.0009367047902468784,
11
+ "std": 0.0
12
+ },
13
+ "multimodality": null,
14
+ "num_samples": 512,
15
+ "r_precision_generated": [
16
+ 0.240234375,
17
+ 0.33984375,
18
+ 0.412109375
19
+ ],
20
+ "r_precision_reference": [
21
+ 0.705078125,
22
+ 0.8359375,
23
+ 0.90234375
24
+ ],
25
+ "text_cache_dir": "outputs/evaluator_text_cache/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f",
26
+ "text_encoder_model": "${OMG_MODELS_ROOT}/text_encoders/t5-3b-local",
27
+ "text_generated": {
28
+ "matching_distance": {
29
+ "max": 1.4769279956817627,
30
+ "mean": 1.331159382360056,
31
+ "min": 1.134010910987854,
32
+ "std": 0.06046604617699041
33
+ },
34
+ "matching_score": 1.3311593532562256,
35
+ "mean_text_rank": 8.705078125,
36
+ "median_text_rank": 6.0,
37
+ "r_precision": [
38
+ 0.240234375,
39
+ 0.33984375,
40
+ 0.412109375
41
+ ],
42
+ "retrieval_batch_size": 32,
43
+ "retrieval_num_batches": 16,
44
+ "retrieval_num_samples": 512,
45
+ "retrieval_order_policy": "stratified_by_dataset",
46
+ "retrieval_unique_datasets_per_batch": {
47
+ "max": 12.0,
48
+ "mean": 12.0,
49
+ "min": 12.0,
50
+ "std": 0.0
51
+ }
52
+ },
53
+ "text_reference": {
54
+ "matching_distance": {
55
+ "max": 1.4135501384735107,
56
+ "mean": 1.2446968206204474,
57
+ "min": 1.1285470724105835,
58
+ "std": 0.0457579048836768
59
+ },
60
+ "matching_score": 1.244696855545044,
61
+ "mean_text_rank": 1.869140625,
62
+ "median_text_rank": 1.0,
63
+ "r_precision": [
64
+ 0.705078125,
65
+ 0.8359375,
66
+ 0.90234375
67
+ ],
68
+ "retrieval_batch_size": 32,
69
+ "retrieval_num_batches": 16,
70
+ "retrieval_num_samples": 512,
71
+ "retrieval_order_policy": "stratified_by_dataset",
72
+ "retrieval_unique_datasets_per_batch": {
73
+ "max": 12.0,
74
+ "mean": 12.0,
75
+ "min": 12.0,
76
+ "std": 0.0
77
+ }
78
+ },
79
+ "text_retrieval_batch_size": 32
80
+ }
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/metrics.json ADDED
@@ -0,0 +1,1922 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark": {
3
+ "benchmark": "artifact_text",
4
+ "dataset_metrics": {
5
+ "100style_original_filtered_test": {
6
+ "embedding": {
7
+ "diversity_generated": 1.0704150905599346,
8
+ "motion_fid": 1.2838937050444919,
9
+ "motion_kid": {
10
+ "mean": 0.0022249471798367892,
11
+ "std": 0.0
12
+ },
13
+ "num_samples": 43
14
+ },
15
+ "num_samples": 43,
16
+ "physical": {
17
+ "body_jerk_mean": {
18
+ "max": 154.36744689941406,
19
+ "mean": 73.70867042208828,
20
+ "min": 2.9786179065704346,
21
+ "std": 37.651289022141334
22
+ },
23
+ "contact_sliding_speed": {
24
+ "max": 1.0541962385177612,
25
+ "mean": 0.3344618064286404,
26
+ "min": 0.007166638039052486,
27
+ "std": 0.2880695016780662
28
+ },
29
+ "foot_ground_error": {
30
+ "max": 0.015996823087334633,
31
+ "mean": 0.005727387244574899,
32
+ "min": 0.0019035614095628262,
33
+ "std": 0.002544252817203053
34
+ },
35
+ "num_samples": 43
36
+ },
37
+ "text": {
38
+ "generated": {
39
+ "matching_distance": {
40
+ "max": 1.4084994792938232,
41
+ "mean": 1.3151017427444458,
42
+ "min": 1.2379783391952515,
43
+ "std": 0.036907895016914875
44
+ },
45
+ "matching_score": 1.3151017427444458,
46
+ "mean_text_rank": 10.90625,
47
+ "median_text_rank": 10.0,
48
+ "r_precision": [
49
+ 0.09375,
50
+ 0.125,
51
+ 0.1875
52
+ ],
53
+ "retrieval_batch_size": 32,
54
+ "retrieval_num_batches": 1,
55
+ "retrieval_num_samples": 32,
56
+ "retrieval_order_policy": "manifest_order",
57
+ "retrieval_unique_datasets_per_batch": {
58
+ "max": 1.0,
59
+ "mean": 1.0,
60
+ "min": 1.0,
61
+ "std": 0.0
62
+ }
63
+ },
64
+ "reference": {
65
+ "matching_distance": {
66
+ "max": 1.3596904277801514,
67
+ "mean": 1.26676583567331,
68
+ "min": 1.1853543519973755,
69
+ "std": 0.03781489479220778
70
+ },
71
+ "matching_score": 1.266765832901001,
72
+ "mean_text_rank": 5.71875,
73
+ "median_text_rank": 3.5,
74
+ "r_precision": [
75
+ 0.34375,
76
+ 0.46875,
77
+ 0.5
78
+ ],
79
+ "retrieval_batch_size": 32,
80
+ "retrieval_num_batches": 1,
81
+ "retrieval_num_samples": 32,
82
+ "retrieval_order_policy": "manifest_order",
83
+ "retrieval_unique_datasets_per_batch": {
84
+ "max": 1.0,
85
+ "mean": 1.0,
86
+ "min": 1.0,
87
+ "std": 0.0
88
+ }
89
+ }
90
+ },
91
+ "transition": {
92
+ "auj": {
93
+ "max": 3.0685779272303186,
94
+ "mean": 1.8053734862221296,
95
+ "min": 1.6700049626518585,
96
+ "std": 0.20044042082439859
97
+ },
98
+ "body_auj": {
99
+ "max": 3.0685779272303186,
100
+ "mean": 1.8053734862221296,
101
+ "min": 1.6700049626518585,
102
+ "std": 0.20044042082439859
103
+ },
104
+ "body_pj": {
105
+ "max": 0.6321938335895538,
106
+ "mean": 0.06189292100756321,
107
+ "min": 0.0008932948112487793,
108
+ "std": 0.09465196402988692
109
+ },
110
+ "boundary_frames": [
111
+ 60
112
+ ],
113
+ "chunk_length": 60,
114
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
115
+ "num_frames": 120,
116
+ "num_samples": 43,
117
+ "pj": {
118
+ "max": 0.6321938335895538,
119
+ "mean": 0.06189292100756321,
120
+ "min": 0.0008932948112487793,
121
+ "std": 0.09465196402988692
122
+ },
123
+ "qpos_auj": {
124
+ "max": 24.71011780718753,
125
+ "mean": 6.1841539223100535,
126
+ "min": 4.2977976540647,
127
+ "std": 3.838152676323723
128
+ },
129
+ "qpos_pj": {
130
+ "max": 9.507055640220642,
131
+ "mean": 0.9185523066323164,
132
+ "min": 0.003028661012649536,
133
+ "std": 1.6940965662424199
134
+ },
135
+ "reference_body_jerk_level": 0.032168975994990016,
136
+ "reference_qpos_jerk_level": 0.07551584170589375,
137
+ "transition_window_length": 60,
138
+ "transition_windows": [
139
+ [
140
+ 30,
141
+ 90
142
+ ]
143
+ ]
144
+ }
145
+ },
146
+ "amass_original_filtered_test": {
147
+ "embedding": {
148
+ "diversity_generated": 1.111925104779673,
149
+ "motion_fid": 1.328561012147938,
150
+ "motion_kid": {
151
+ "mean": 0.0022260718776268007,
152
+ "std": 0.0
153
+ },
154
+ "num_samples": 43
155
+ },
156
+ "num_samples": 43,
157
+ "physical": {
158
+ "body_jerk_mean": {
159
+ "max": 269.9172668457031,
160
+ "mean": 95.9706466918768,
161
+ "min": 2.5841948986053467,
162
+ "std": 70.75177739661642
163
+ },
164
+ "contact_sliding_speed": {
165
+ "max": 1.5415642261505127,
166
+ "mean": 0.39597279582779077,
167
+ "min": 0.005511525087058544,
168
+ "std": 0.40617249845569314
169
+ },
170
+ "foot_ground_error": {
171
+ "max": 0.04935852810740471,
172
+ "mean": 0.00705925366559694,
173
+ "min": 0.0024636401794850826,
174
+ "std": 0.007562361206799781
175
+ },
176
+ "num_samples": 43
177
+ },
178
+ "text": {
179
+ "generated": {
180
+ "matching_distance": {
181
+ "max": 1.418434739112854,
182
+ "mean": 1.3283776654753574,
183
+ "min": 1.2401793003082275,
184
+ "std": 0.04136207243662587
185
+ },
186
+ "matching_score": 1.328377604484558,
187
+ "mean_text_rank": 6.96875,
188
+ "median_text_rank": 3.0,
189
+ "r_precision": [
190
+ 0.375,
191
+ 0.5,
192
+ 0.5
193
+ ],
194
+ "retrieval_batch_size": 32,
195
+ "retrieval_num_batches": 1,
196
+ "retrieval_num_samples": 32,
197
+ "retrieval_order_policy": "manifest_order",
198
+ "retrieval_unique_datasets_per_batch": {
199
+ "max": 1.0,
200
+ "mean": 1.0,
201
+ "min": 1.0,
202
+ "std": 0.0
203
+ }
204
+ },
205
+ "reference": {
206
+ "matching_distance": {
207
+ "max": 1.4135501384735107,
208
+ "mean": 1.272889874702276,
209
+ "min": 1.172876000404358,
210
+ "std": 0.04160945025899305
211
+ },
212
+ "matching_score": 1.2728898525238037,
213
+ "mean_text_rank": 4.40625,
214
+ "median_text_rank": 2.0,
215
+ "r_precision": [
216
+ 0.3125,
217
+ 0.5625,
218
+ 0.625
219
+ ],
220
+ "retrieval_batch_size": 32,
221
+ "retrieval_num_batches": 1,
222
+ "retrieval_num_samples": 32,
223
+ "retrieval_order_policy": "manifest_order",
224
+ "retrieval_unique_datasets_per_batch": {
225
+ "max": 1.0,
226
+ "mean": 1.0,
227
+ "min": 1.0,
228
+ "std": 0.0
229
+ }
230
+ }
231
+ },
232
+ "transition": {
233
+ "auj": {
234
+ "max": 6.514222457101631,
235
+ "mean": 1.8822794797577118,
236
+ "min": 1.5973546691528981,
237
+ "std": 0.7210500357139227
238
+ },
239
+ "body_auj": {
240
+ "max": 6.514222457101631,
241
+ "mean": 1.8822794797577118,
242
+ "min": 1.5973546691528981,
243
+ "std": 0.7210500357139227
244
+ },
245
+ "body_pj": {
246
+ "max": 1.9185047969222069,
247
+ "mean": 0.09840007811828061,
248
+ "min": 0.00058785080909729,
249
+ "std": 0.28364867706203306
250
+ },
251
+ "boundary_frames": [
252
+ 60
253
+ ],
254
+ "chunk_length": 60,
255
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
256
+ "num_frames": 120,
257
+ "num_samples": 43,
258
+ "pj": {
259
+ "max": 1.9185047969222069,
260
+ "mean": 0.09840007811828061,
261
+ "min": 0.00058785080909729,
262
+ "std": 0.28364867706203306
263
+ },
264
+ "qpos_auj": {
265
+ "max": 72.83266460476783,
266
+ "mean": 8.4905600420132,
267
+ "min": 4.2940439462623115,
268
+ "std": 11.744696961776553
269
+ },
270
+ "qpos_pj": {
271
+ "max": 12.584792256355286,
272
+ "mean": 1.159780290310791,
273
+ "min": 0.002947501838207245,
274
+ "std": 2.334767403530377
275
+ },
276
+ "reference_body_jerk_level": 0.032168975994990016,
277
+ "reference_qpos_jerk_level": 0.07551584170589375,
278
+ "transition_window_length": 60,
279
+ "transition_windows": [
280
+ [
281
+ 30,
282
+ 90
283
+ ]
284
+ ]
285
+ }
286
+ },
287
+ "bones_seed_original_filtered_test": {
288
+ "embedding": {
289
+ "diversity_generated": 1.1937374883617156,
290
+ "motion_fid": 0.4045952532270245,
291
+ "motion_kid": {
292
+ "mean": -0.00010609229304492374,
293
+ "std": 0.0
294
+ },
295
+ "num_samples": 43
296
+ },
297
+ "num_samples": 43,
298
+ "physical": {
299
+ "body_jerk_mean": {
300
+ "max": 261.4084777832031,
301
+ "mean": 69.5281113635662,
302
+ "min": 3.05818772315979,
303
+ "std": 68.85240862628845
304
+ },
305
+ "contact_sliding_speed": {
306
+ "max": 2.0296926498413086,
307
+ "mean": 0.5391670568348017,
308
+ "min": 0.004975942894816399,
309
+ "std": 0.5557337065044728
310
+ },
311
+ "foot_ground_error": {
312
+ "max": 0.012411188334226608,
313
+ "mean": 0.0062796306758469275,
314
+ "min": 0.0013083467492833734,
315
+ "std": 0.0025643484083098323
316
+ },
317
+ "num_samples": 43
318
+ },
319
+ "text": {
320
+ "generated": {
321
+ "matching_distance": {
322
+ "max": 1.310302972793579,
323
+ "mean": 1.2274910932363465,
324
+ "min": 1.134010910987854,
325
+ "std": 0.03699650409915577
326
+ },
327
+ "matching_score": 1.2274911403656006,
328
+ "mean_text_rank": 1.5,
329
+ "median_text_rank": 1.0,
330
+ "r_precision": [
331
+ 0.71875,
332
+ 0.90625,
333
+ 0.9375
334
+ ],
335
+ "retrieval_batch_size": 32,
336
+ "retrieval_num_batches": 1,
337
+ "retrieval_num_samples": 32,
338
+ "retrieval_order_policy": "manifest_order",
339
+ "retrieval_unique_datasets_per_batch": {
340
+ "max": 1.0,
341
+ "mean": 1.0,
342
+ "min": 1.0,
343
+ "std": 0.0
344
+ }
345
+ },
346
+ "reference": {
347
+ "matching_distance": {
348
+ "max": 1.3188464641571045,
349
+ "mean": 1.2265774760135384,
350
+ "min": 1.1549649238586426,
351
+ "std": 0.04007899902186925
352
+ },
353
+ "matching_score": 1.2265775203704834,
354
+ "mean_text_rank": 1.84375,
355
+ "median_text_rank": 1.0,
356
+ "r_precision": [
357
+ 0.6875,
358
+ 0.84375,
359
+ 0.90625
360
+ ],
361
+ "retrieval_batch_size": 32,
362
+ "retrieval_num_batches": 1,
363
+ "retrieval_num_samples": 32,
364
+ "retrieval_order_policy": "manifest_order",
365
+ "retrieval_unique_datasets_per_batch": {
366
+ "max": 1.0,
367
+ "mean": 1.0,
368
+ "min": 1.0,
369
+ "std": 0.0
370
+ }
371
+ }
372
+ },
373
+ "transition": {
374
+ "auj": {
375
+ "max": 1.9905637383757522,
376
+ "mean": 1.7810514533823214,
377
+ "min": 1.5876606825310595,
378
+ "std": 0.06310559905267207
379
+ },
380
+ "body_auj": {
381
+ "max": 1.9905637383757522,
382
+ "mean": 1.7810514533823214,
383
+ "min": 1.5876606825310595,
384
+ "std": 0.06310559905267207
385
+ },
386
+ "body_pj": {
387
+ "max": 0.12257324159145355,
388
+ "mean": 0.04556534072210969,
389
+ "min": 0.0018733292818069458,
390
+ "std": 0.03456947191402774
391
+ },
392
+ "boundary_frames": [
393
+ 60
394
+ ],
395
+ "chunk_length": 60,
396
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
397
+ "num_frames": 120,
398
+ "num_samples": 43,
399
+ "pj": {
400
+ "max": 0.12257324159145355,
401
+ "mean": 0.04556534072210969,
402
+ "min": 0.0018733292818069458,
403
+ "std": 0.03456947191402774
404
+ },
405
+ "qpos_auj": {
406
+ "max": 20.530484854076903,
407
+ "mean": 6.420968679559986,
408
+ "min": 4.301883842105374,
409
+ "std": 4.107536056696693
410
+ },
411
+ "qpos_pj": {
412
+ "max": 6.2786277532577515,
413
+ "mean": 0.8628149387530907,
414
+ "min": 0.019462376832962036,
415
+ "std": 1.4660741357554568
416
+ },
417
+ "reference_body_jerk_level": 0.032168975994990016,
418
+ "reference_qpos_jerk_level": 0.07551584170589375,
419
+ "transition_window_length": 60,
420
+ "transition_windows": [
421
+ [
422
+ 30,
423
+ 90
424
+ ]
425
+ ]
426
+ }
427
+ },
428
+ "fitness_original_filtered_test": {
429
+ "embedding": {
430
+ "diversity_generated": 1.1245486499453385,
431
+ "motion_fid": 1.3721207787485414,
432
+ "motion_kid": {
433
+ "mean": 0.0019768632581005363,
434
+ "std": 0.0
435
+ },
436
+ "num_samples": 43
437
+ },
438
+ "num_samples": 43,
439
+ "physical": {
440
+ "body_jerk_mean": {
441
+ "max": 307.0351257324219,
442
+ "mean": 62.3401137174562,
443
+ "min": 5.94058895111084,
444
+ "std": 74.72415873290508
445
+ },
446
+ "contact_sliding_speed": {
447
+ "max": 1.9512882232666016,
448
+ "mean": 0.20101934357351342,
449
+ "min": 0.007332059554755688,
450
+ "std": 0.4436527359467069
451
+ },
452
+ "foot_ground_error": {
453
+ "max": 0.012699918821454048,
454
+ "mean": 0.006037447639459441,
455
+ "min": 0.0020615984685719013,
456
+ "std": 0.0021288537281194384
457
+ },
458
+ "num_samples": 43
459
+ },
460
+ "text": {
461
+ "generated": {
462
+ "matching_distance": {
463
+ "max": 1.460925579071045,
464
+ "mean": 1.3796154177466105,
465
+ "min": 1.268312692642212,
466
+ "std": 0.04377553385148069
467
+ },
468
+ "matching_score": 1.3796154260635376,
469
+ "mean_text_rank": 10.46875,
470
+ "median_text_rank": 11.0,
471
+ "r_precision": [
472
+ 0.125,
473
+ 0.125,
474
+ 0.125
475
+ ],
476
+ "retrieval_batch_size": 32,
477
+ "retrieval_num_batches": 1,
478
+ "retrieval_num_samples": 32,
479
+ "retrieval_order_policy": "manifest_order",
480
+ "retrieval_unique_datasets_per_batch": {
481
+ "max": 1.0,
482
+ "mean": 1.0,
483
+ "min": 1.0,
484
+ "std": 0.0
485
+ }
486
+ },
487
+ "reference": {
488
+ "matching_distance": {
489
+ "max": 1.3819284439086914,
490
+ "mean": 1.2803652868714444,
491
+ "min": 1.209172010421753,
492
+ "std": 0.03319129105747331
493
+ },
494
+ "matching_score": 1.280365228652954,
495
+ "mean_text_rank": 7.5,
496
+ "median_text_rank": 7.5,
497
+ "r_precision": [
498
+ 0.21875,
499
+ 0.25,
500
+ 0.3125
501
+ ],
502
+ "retrieval_batch_size": 32,
503
+ "retrieval_num_batches": 1,
504
+ "retrieval_num_samples": 32,
505
+ "retrieval_order_policy": "manifest_order",
506
+ "retrieval_unique_datasets_per_batch": {
507
+ "max": 1.0,
508
+ "mean": 1.0,
509
+ "min": 1.0,
510
+ "std": 0.0
511
+ }
512
+ }
513
+ },
514
+ "transition": {
515
+ "auj": {
516
+ "max": 2.385969114632759,
517
+ "mean": 1.8030387962130037,
518
+ "min": 1.4960776350777631,
519
+ "std": 0.11265973690452995
520
+ },
521
+ "body_auj": {
522
+ "max": 2.385969114632759,
523
+ "mean": 1.8030387962130037,
524
+ "min": 1.4960776350777631,
525
+ "std": 0.11265973690452995
526
+ },
527
+ "body_pj": {
528
+ "max": 0.21083378558978438,
529
+ "mean": 0.04393309076468265,
530
+ "min": 0.0029095709323883057,
531
+ "std": 0.0441904998385772
532
+ },
533
+ "boundary_frames": [
534
+ 60
535
+ ],
536
+ "chunk_length": 60,
537
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
538
+ "num_frames": 120,
539
+ "num_samples": 43,
540
+ "pj": {
541
+ "max": 0.21083378558978438,
542
+ "mean": 0.04393309076468265,
543
+ "min": 0.0029095709323883057,
544
+ "std": 0.0441904998385772
545
+ },
546
+ "qpos_auj": {
547
+ "max": 144.49321365365512,
548
+ "mean": 7.995620672684568,
549
+ "min": 4.299943787984878,
550
+ "std": 21.08138489740843
551
+ },
552
+ "qpos_pj": {
553
+ "max": 12.581249475479126,
554
+ "mean": 0.4969454543586539,
555
+ "min": 0.011551141738891602,
556
+ "std": 1.8728602685043632
557
+ },
558
+ "reference_body_jerk_level": 0.032168975994990016,
559
+ "reference_qpos_jerk_level": 0.07551584170589375,
560
+ "transition_window_length": 60,
561
+ "transition_windows": [
562
+ [
563
+ 30,
564
+ 90
565
+ ]
566
+ ]
567
+ }
568
+ },
569
+ "humanml_original_filtered_test": {
570
+ "embedding": {
571
+ "diversity_generated": 1.1726888731015561,
572
+ "motion_fid": 1.175771434491134,
573
+ "motion_kid": {
574
+ "mean": 0.0014073873525246583,
575
+ "std": 0.0
576
+ },
577
+ "num_samples": 43
578
+ },
579
+ "num_samples": 43,
580
+ "physical": {
581
+ "body_jerk_mean": {
582
+ "max": 246.43716430664062,
583
+ "mean": 68.01800821548285,
584
+ "min": 2.1355345249176025,
585
+ "std": 57.236756620842854
586
+ },
587
+ "contact_sliding_speed": {
588
+ "max": 1.9895073175430298,
589
+ "mean": 0.39400377310812473,
590
+ "min": 0.0,
591
+ "std": 0.46918049539700873
592
+ },
593
+ "foot_ground_error": {
594
+ "max": 1.477441668510437,
595
+ "mean": 0.04231151855594024,
596
+ "min": 0.002110379049554467,
597
+ "std": 0.2215764948289639
598
+ },
599
+ "num_samples": 43
600
+ },
601
+ "text": {
602
+ "generated": {
603
+ "matching_distance": {
604
+ "max": 1.430628776550293,
605
+ "mean": 1.3250223120977713,
606
+ "min": 1.2387126684188843,
607
+ "std": 0.04734495346885312
608
+ },
609
+ "matching_score": 1.3250223398208618,
610
+ "mean_text_rank": 6.71875,
611
+ "median_text_rank": 3.0,
612
+ "r_precision": [
613
+ 0.28125,
614
+ 0.46875,
615
+ 0.5625
616
+ ],
617
+ "retrieval_batch_size": 32,
618
+ "retrieval_num_batches": 1,
619
+ "retrieval_num_samples": 32,
620
+ "retrieval_order_policy": "manifest_order",
621
+ "retrieval_unique_datasets_per_batch": {
622
+ "max": 1.0,
623
+ "mean": 1.0,
624
+ "min": 1.0,
625
+ "std": 0.0
626
+ }
627
+ },
628
+ "reference": {
629
+ "matching_distance": {
630
+ "max": 1.3049654960632324,
631
+ "mean": 1.2212754654329876,
632
+ "min": 1.151880145072937,
633
+ "std": 0.033672752576929775
634
+ },
635
+ "matching_score": 1.2212754487991333,
636
+ "mean_text_rank": 2.0625,
637
+ "median_text_rank": 1.0,
638
+ "r_precision": [
639
+ 0.53125,
640
+ 0.78125,
641
+ 0.875
642
+ ],
643
+ "retrieval_batch_size": 32,
644
+ "retrieval_num_batches": 1,
645
+ "retrieval_num_samples": 32,
646
+ "retrieval_order_policy": "manifest_order",
647
+ "retrieval_unique_datasets_per_batch": {
648
+ "max": 1.0,
649
+ "mean": 1.0,
650
+ "min": 1.0,
651
+ "std": 0.0
652
+ }
653
+ }
654
+ },
655
+ "transition": {
656
+ "auj": {
657
+ "max": 2.2740369624934447,
658
+ "mean": 1.7931604724291397,
659
+ "min": 1.6542664254946329,
660
+ "std": 0.08686737366358381
661
+ },
662
+ "body_auj": {
663
+ "max": 2.2740369624934447,
664
+ "mean": 1.7931604724291397,
665
+ "min": 1.6542664254946329,
666
+ "std": 0.08686737366358381
667
+ },
668
+ "body_pj": {
669
+ "max": 0.21533825621008873,
670
+ "mean": 0.04874728411046225,
671
+ "min": 0.0004047006368637085,
672
+ "std": 0.04271094174830256
673
+ },
674
+ "boundary_frames": [
675
+ 60
676
+ ],
677
+ "chunk_length": 60,
678
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
679
+ "num_frames": 120,
680
+ "num_samples": 43,
681
+ "pj": {
682
+ "max": 0.21533825621008873,
683
+ "mean": 0.04874728411046225,
684
+ "min": 0.0004047006368637085,
685
+ "std": 0.04271094174830256
686
+ },
687
+ "qpos_auj": {
688
+ "max": 10.246992179070007,
689
+ "mean": 5.0103711591434354,
690
+ "min": 4.300404030837447,
691
+ "std": 1.3281499654605524
692
+ },
693
+ "qpos_pj": {
694
+ "max": 2.85389244556427,
695
+ "mean": 0.3255547105367187,
696
+ "min": 0.0028911828994750977,
697
+ "std": 0.5239349095283596
698
+ },
699
+ "reference_body_jerk_level": 0.032168975994990016,
700
+ "reference_qpos_jerk_level": 0.07551584170589375,
701
+ "transition_window_length": 60,
702
+ "transition_windows": [
703
+ [
704
+ 30,
705
+ 90
706
+ ]
707
+ ]
708
+ }
709
+ },
710
+ "idea400_original_filtered_test": {
711
+ "embedding": {
712
+ "diversity_generated": 1.1694871863472327,
713
+ "motion_fid": 1.424413235305793,
714
+ "motion_kid": {
715
+ "mean": 0.003020059186301882,
716
+ "std": 0.0
717
+ },
718
+ "num_samples": 43
719
+ },
720
+ "num_samples": 43,
721
+ "physical": {
722
+ "body_jerk_mean": {
723
+ "max": 185.40188598632812,
724
+ "mean": 48.86681900467983,
725
+ "min": 2.1356678009033203,
726
+ "std": 52.217468326814576
727
+ },
728
+ "contact_sliding_speed": {
729
+ "max": 1.2346510887145996,
730
+ "mean": 0.27460066539365363,
731
+ "min": 0.0036743313539773226,
732
+ "std": 0.3540692147671962
733
+ },
734
+ "foot_ground_error": {
735
+ "max": 0.01912386529147625,
736
+ "mean": 0.0067057522327834095,
737
+ "min": 0.00249672238714993,
738
+ "std": 0.003613244239824124
739
+ },
740
+ "num_samples": 43
741
+ },
742
+ "text": {
743
+ "generated": {
744
+ "matching_distance": {
745
+ "max": 1.4311089515686035,
746
+ "mean": 1.3534246184105096,
747
+ "min": 1.2549099922180176,
748
+ "std": 0.040087949446493104
749
+ },
750
+ "matching_score": 1.3534245491027832,
751
+ "mean_text_rank": 8.5,
752
+ "median_text_rank": 5.0,
753
+ "r_precision": [
754
+ 0.1875,
755
+ 0.21875,
756
+ 0.3125
757
+ ],
758
+ "retrieval_batch_size": 32,
759
+ "retrieval_num_batches": 1,
760
+ "retrieval_num_samples": 32,
761
+ "retrieval_order_policy": "manifest_order",
762
+ "retrieval_unique_datasets_per_batch": {
763
+ "max": 1.0,
764
+ "mean": 1.0,
765
+ "min": 1.0,
766
+ "std": 0.0
767
+ }
768
+ },
769
+ "reference": {
770
+ "matching_distance": {
771
+ "max": 1.3933742046356201,
772
+ "mean": 1.253553493078365,
773
+ "min": 1.1590988636016846,
774
+ "std": 0.0395202842188391
775
+ },
776
+ "matching_score": 1.2535533905029297,
777
+ "mean_text_rank": 5.5625,
778
+ "median_text_rank": 4.0,
779
+ "r_precision": [
780
+ 0.21875,
781
+ 0.4375,
782
+ 0.46875
783
+ ],
784
+ "retrieval_batch_size": 32,
785
+ "retrieval_num_batches": 1,
786
+ "retrieval_num_samples": 32,
787
+ "retrieval_order_policy": "manifest_order",
788
+ "retrieval_unique_datasets_per_batch": {
789
+ "max": 1.0,
790
+ "mean": 1.0,
791
+ "min": 1.0,
792
+ "std": 0.0
793
+ }
794
+ }
795
+ },
796
+ "transition": {
797
+ "auj": {
798
+ "max": 1.9275732800142191,
799
+ "mean": 1.7972521730343467,
800
+ "min": 1.582865984803458,
801
+ "std": 0.06215447215088484
802
+ },
803
+ "body_auj": {
804
+ "max": 1.9275732800142191,
805
+ "mean": 1.7972521730343467,
806
+ "min": 1.582865984803458,
807
+ "std": 0.06215447215088484
808
+ },
809
+ "body_pj": {
810
+ "max": 0.19542989134788513,
811
+ "mean": 0.037036049071439474,
812
+ "min": 0.00043263286352157593,
813
+ "std": 0.04102339681952739
814
+ },
815
+ "boundary_frames": [
816
+ 60
817
+ ],
818
+ "chunk_length": 60,
819
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
820
+ "num_frames": 120,
821
+ "num_samples": 43,
822
+ "pj": {
823
+ "max": 0.19542989134788513,
824
+ "mean": 0.037036049071439474,
825
+ "min": 0.00043263286352157593,
826
+ "std": 0.04102339681952739
827
+ },
828
+ "qpos_auj": {
829
+ "max": 35.02763120426745,
830
+ "mean": 5.4219513973630304,
831
+ "min": 4.3006585100743475,
832
+ "std": 4.6576733588000225
833
+ },
834
+ "qpos_pj": {
835
+ "max": 12.468976974487305,
836
+ "mean": 0.5249752198849523,
837
+ "min": 0.002636820077896118,
838
+ "std": 1.8992311316861998
839
+ },
840
+ "reference_body_jerk_level": 0.032168975994990016,
841
+ "reference_qpos_jerk_level": 0.07551584170589375,
842
+ "transition_window_length": 60,
843
+ "transition_windows": [
844
+ [
845
+ 30,
846
+ 90
847
+ ]
848
+ ]
849
+ }
850
+ },
851
+ "lafan1_original_filtered_test": {
852
+ "embedding": {
853
+ "diversity_generated": 1.065131755766148,
854
+ "motion_fid": 0.9163817643240386,
855
+ "motion_kid": {
856
+ "mean": 0.001431817572137728,
857
+ "std": 0.0
858
+ },
859
+ "num_samples": 43
860
+ },
861
+ "num_samples": 43,
862
+ "physical": {
863
+ "body_jerk_mean": {
864
+ "max": 531.6823120117188,
865
+ "mean": 164.51493027043898,
866
+ "min": 6.81342077255249,
867
+ "std": 135.10997614115757
868
+ },
869
+ "contact_sliding_speed": {
870
+ "max": 2.914012908935547,
871
+ "mean": 0.8355859869354686,
872
+ "min": 0.008567381650209427,
873
+ "std": 0.7517330338413282
874
+ },
875
+ "foot_ground_error": {
876
+ "max": 0.013226441107690334,
877
+ "mean": 0.0067459214966051105,
878
+ "min": 0.0020911949686706066,
879
+ "std": 0.002395367031927562
880
+ },
881
+ "num_samples": 43
882
+ },
883
+ "text": {
884
+ "generated": {
885
+ "matching_distance": {
886
+ "max": 1.3756978511810303,
887
+ "mean": 1.2848902968473213,
888
+ "min": 1.22147798538208,
889
+ "std": 0.03506700337563933
890
+ },
891
+ "matching_score": 1.2848902940750122,
892
+ "mean_text_rank": 6.78125,
893
+ "median_text_rank": 5.0,
894
+ "r_precision": [
895
+ 0.375,
896
+ 0.40625,
897
+ 0.46875
898
+ ],
899
+ "retrieval_batch_size": 32,
900
+ "retrieval_num_batches": 1,
901
+ "retrieval_num_samples": 32,
902
+ "retrieval_order_policy": "manifest_order",
903
+ "retrieval_unique_datasets_per_batch": {
904
+ "max": 1.0,
905
+ "mean": 1.0,
906
+ "min": 1.0,
907
+ "std": 0.0
908
+ }
909
+ },
910
+ "reference": {
911
+ "matching_distance": {
912
+ "max": 1.3601514101028442,
913
+ "mean": 1.262607499610546,
914
+ "min": 1.1779136657714844,
915
+ "std": 0.036237216916877986
916
+ },
917
+ "matching_score": 1.262607455253601,
918
+ "mean_text_rank": 7.34375,
919
+ "median_text_rank": 4.0,
920
+ "r_precision": [
921
+ 0.28125,
922
+ 0.34375,
923
+ 0.46875
924
+ ],
925
+ "retrieval_batch_size": 32,
926
+ "retrieval_num_batches": 1,
927
+ "retrieval_num_samples": 32,
928
+ "retrieval_order_policy": "manifest_order",
929
+ "retrieval_unique_datasets_per_batch": {
930
+ "max": 1.0,
931
+ "mean": 1.0,
932
+ "min": 1.0,
933
+ "std": 0.0
934
+ }
935
+ }
936
+ },
937
+ "transition": {
938
+ "auj": {
939
+ "max": 4.683557439704222,
940
+ "mean": 1.9927798923981312,
941
+ "min": 1.7141404278200993,
942
+ "std": 0.5034205327935629
943
+ },
944
+ "body_auj": {
945
+ "max": 4.683557439704222,
946
+ "mean": 1.9927798923981312,
947
+ "min": 1.7141404278200993,
948
+ "std": 0.5034205327935629
949
+ },
950
+ "body_pj": {
951
+ "max": 0.8904187381267548,
952
+ "mean": 0.11073192904248487,
953
+ "min": 0.0009953081607818604,
954
+ "std": 0.13824281051787887
955
+ },
956
+ "boundary_frames": [
957
+ 60
958
+ ],
959
+ "chunk_length": 60,
960
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
961
+ "num_frames": 120,
962
+ "num_samples": 43,
963
+ "pj": {
964
+ "max": 0.8904187381267548,
965
+ "mean": 0.11073192904248487,
966
+ "min": 0.0009953081607818604,
967
+ "std": 0.13824281051787887
968
+ },
969
+ "qpos_auj": {
970
+ "max": 43.07620444184711,
971
+ "mean": 8.566258157005608,
972
+ "min": 4.297486459604036,
973
+ "std": 7.636360765758662
974
+ },
975
+ "qpos_pj": {
976
+ "max": 12.626699447631836,
977
+ "mean": 1.2561438795352398,
978
+ "min": 0.00521540641784668,
979
+ "std": 2.344095482079285
980
+ },
981
+ "reference_body_jerk_level": 0.032168975994990016,
982
+ "reference_qpos_jerk_level": 0.07551584170589375,
983
+ "transition_window_length": 60,
984
+ "transition_windows": [
985
+ [
986
+ 30,
987
+ 90
988
+ ]
989
+ ]
990
+ }
991
+ },
992
+ "motiongv_original_filtered_test": {
993
+ "embedding": {
994
+ "diversity_generated": 1.151175892575208,
995
+ "motion_fid": 1.4664699422882885,
996
+ "motion_kid": {
997
+ "mean": 0.0018693371237699097,
998
+ "std": 0.0
999
+ },
1000
+ "num_samples": 43
1001
+ },
1002
+ "num_samples": 43,
1003
+ "physical": {
1004
+ "body_jerk_mean": {
1005
+ "max": 391.6789245605469,
1006
+ "mean": 82.85689859612043,
1007
+ "min": 3.0611748695373535,
1008
+ "std": 95.17589490486428
1009
+ },
1010
+ "contact_sliding_speed": {
1011
+ "max": 1.5828344821929932,
1012
+ "mean": 0.17748946919604097,
1013
+ "min": 0.006206035614013672,
1014
+ "std": 0.2848895833067525
1015
+ },
1016
+ "foot_ground_error": {
1017
+ "max": 0.05884654074907303,
1018
+ "mean": 0.00999552955266175,
1019
+ "min": 0.0020414358004927635,
1020
+ "std": 0.009597411811076375
1021
+ },
1022
+ "num_samples": 43
1023
+ },
1024
+ "text": {
1025
+ "generated": {
1026
+ "matching_distance": {
1027
+ "max": 1.472267508506775,
1028
+ "mean": 1.3855077416397805,
1029
+ "min": 1.2616697549819946,
1030
+ "std": 0.04268846702374364
1031
+ },
1032
+ "matching_score": 1.3855078220367432,
1033
+ "mean_text_rank": 6.71875,
1034
+ "median_text_rank": 3.5,
1035
+ "r_precision": [
1036
+ 0.28125,
1037
+ 0.40625,
1038
+ 0.5
1039
+ ],
1040
+ "retrieval_batch_size": 32,
1041
+ "retrieval_num_batches": 1,
1042
+ "retrieval_num_samples": 32,
1043
+ "retrieval_order_policy": "manifest_order",
1044
+ "retrieval_unique_datasets_per_batch": {
1045
+ "max": 1.0,
1046
+ "mean": 1.0,
1047
+ "min": 1.0,
1048
+ "std": 0.0
1049
+ }
1050
+ },
1051
+ "reference": {
1052
+ "matching_distance": {
1053
+ "max": 1.355556845664978,
1054
+ "mean": 1.2528588660927706,
1055
+ "min": 1.1668779850006104,
1056
+ "std": 0.04499687839541838
1057
+ },
1058
+ "matching_score": 1.2528588771820068,
1059
+ "mean_text_rank": 1.9375,
1060
+ "median_text_rank": 1.0,
1061
+ "r_precision": [
1062
+ 0.65625,
1063
+ 0.8125,
1064
+ 0.8125
1065
+ ],
1066
+ "retrieval_batch_size": 32,
1067
+ "retrieval_num_batches": 1,
1068
+ "retrieval_num_samples": 32,
1069
+ "retrieval_order_policy": "manifest_order",
1070
+ "retrieval_unique_datasets_per_batch": {
1071
+ "max": 1.0,
1072
+ "mean": 1.0,
1073
+ "min": 1.0,
1074
+ "std": 0.0
1075
+ }
1076
+ }
1077
+ },
1078
+ "transition": {
1079
+ "auj": {
1080
+ "max": 3.7732563196112654,
1081
+ "mean": 1.877907262608954,
1082
+ "min": 1.4860971376024386,
1083
+ "std": 0.3515215707606796
1084
+ },
1085
+ "body_auj": {
1086
+ "max": 3.7732563196112654,
1087
+ "mean": 1.877907262608954,
1088
+ "min": 1.4860971376024386,
1089
+ "std": 0.3515215707606796
1090
+ },
1091
+ "body_pj": {
1092
+ "max": 0.5607699304819107,
1093
+ "mean": 0.059313230828360416,
1094
+ "min": 0.0006179101765155792,
1095
+ "std": 0.09926593614052223
1096
+ },
1097
+ "boundary_frames": [
1098
+ 60
1099
+ ],
1100
+ "chunk_length": 60,
1101
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1102
+ "num_frames": 120,
1103
+ "num_samples": 43,
1104
+ "pj": {
1105
+ "max": 0.5607699304819107,
1106
+ "mean": 0.059313230828360416,
1107
+ "min": 0.0006179101765155792,
1108
+ "std": 0.09926593614052223
1109
+ },
1110
+ "qpos_auj": {
1111
+ "max": 62.14833038614482,
1112
+ "mean": 7.156893734841608,
1113
+ "min": 4.296862431809634,
1114
+ "std": 10.065915962574447
1115
+ },
1116
+ "qpos_pj": {
1117
+ "max": 12.888956546783447,
1118
+ "mean": 0.8218644026580245,
1119
+ "min": 0.003064531832933426,
1120
+ "std": 2.6184541571050173
1121
+ },
1122
+ "reference_body_jerk_level": 0.032168975994990016,
1123
+ "reference_qpos_jerk_level": 0.07551584170589375,
1124
+ "transition_window_length": 60,
1125
+ "transition_windows": [
1126
+ [
1127
+ 30,
1128
+ 90
1129
+ ]
1130
+ ]
1131
+ }
1132
+ },
1133
+ "motionllama_original_filtered_test": {
1134
+ "embedding": {
1135
+ "diversity_generated": 1.1930282606242615,
1136
+ "motion_fid": 1.3093164078052215,
1137
+ "motion_kid": {
1138
+ "mean": 0.001977269528855352,
1139
+ "std": 0.0
1140
+ },
1141
+ "num_samples": 42
1142
+ },
1143
+ "num_samples": 42,
1144
+ "physical": {
1145
+ "body_jerk_mean": {
1146
+ "max": 185.41676330566406,
1147
+ "mean": 17.548739166486833,
1148
+ "min": 2.4985158443450928,
1149
+ "std": 29.38218129714949
1150
+ },
1151
+ "contact_sliding_speed": {
1152
+ "max": 1.210018515586853,
1153
+ "mean": 0.05310293933421019,
1154
+ "min": 0.004471967928111553,
1155
+ "std": 0.18190039828460755
1156
+ },
1157
+ "foot_ground_error": {
1158
+ "max": 0.11203083395957947,
1159
+ "mean": 0.008613292068254114,
1160
+ "min": 0.0008450289606116712,
1161
+ "std": 0.016888853227158128
1162
+ },
1163
+ "num_samples": 42
1164
+ },
1165
+ "text": {
1166
+ "generated": {
1167
+ "matching_distance": {
1168
+ "max": 1.4584448337554932,
1169
+ "mean": 1.3611151774724324,
1170
+ "min": 1.2479358911514282,
1171
+ "std": 0.0521328721850203
1172
+ },
1173
+ "matching_score": 1.3611150979995728,
1174
+ "mean_text_rank": 7.75,
1175
+ "median_text_rank": 4.0,
1176
+ "r_precision": [
1177
+ 0.1875,
1178
+ 0.34375,
1179
+ 0.46875
1180
+ ],
1181
+ "retrieval_batch_size": 32,
1182
+ "retrieval_num_batches": 1,
1183
+ "retrieval_num_samples": 32,
1184
+ "retrieval_order_policy": "manifest_order",
1185
+ "retrieval_unique_datasets_per_batch": {
1186
+ "max": 1.0,
1187
+ "mean": 1.0,
1188
+ "min": 1.0,
1189
+ "std": 0.0
1190
+ }
1191
+ },
1192
+ "reference": {
1193
+ "matching_distance": {
1194
+ "max": 1.344225287437439,
1195
+ "mean": 1.2109532526561193,
1196
+ "min": 1.1528350114822388,
1197
+ "std": 0.03675615838449094
1198
+ },
1199
+ "matching_score": 1.2109531164169312,
1200
+ "mean_text_rank": 2.40625,
1201
+ "median_text_rank": 1.0,
1202
+ "r_precision": [
1203
+ 0.53125,
1204
+ 0.6875,
1205
+ 0.8125
1206
+ ],
1207
+ "retrieval_batch_size": 32,
1208
+ "retrieval_num_batches": 1,
1209
+ "retrieval_num_samples": 32,
1210
+ "retrieval_order_policy": "manifest_order",
1211
+ "retrieval_unique_datasets_per_batch": {
1212
+ "max": 1.0,
1213
+ "mean": 1.0,
1214
+ "min": 1.0,
1215
+ "std": 0.0
1216
+ }
1217
+ }
1218
+ },
1219
+ "transition": {
1220
+ "auj": {
1221
+ "max": 1.8308297701004421,
1222
+ "mean": 1.8203518604107993,
1223
+ "min": 1.7468128703884296,
1224
+ "std": 0.015218604769691082
1225
+ },
1226
+ "body_auj": {
1227
+ "max": 1.8308297701004421,
1228
+ "mean": 1.8203518604107993,
1229
+ "min": 1.7468128703884296,
1230
+ "std": 0.015218604769691082
1231
+ },
1232
+ "body_pj": {
1233
+ "max": 0.09689441323280334,
1234
+ "mean": 0.011014723008321155,
1235
+ "min": 0.0005841478705406189,
1236
+ "std": 0.017570536124353737
1237
+ },
1238
+ "boundary_frames": [
1239
+ 60
1240
+ ],
1241
+ "chunk_length": 60,
1242
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1243
+ "num_frames": 120,
1244
+ "num_samples": 42,
1245
+ "pj": {
1246
+ "max": 0.09689441323280334,
1247
+ "mean": 0.011014723008321155,
1248
+ "min": 0.0005841478705406189,
1249
+ "std": 0.017570536124353737
1250
+ },
1251
+ "qpos_auj": {
1252
+ "max": 9.52280969330879,
1253
+ "mean": 4.517935173208866,
1254
+ "min": 4.3012650981623235,
1255
+ "std": 0.8728498919105634
1256
+ },
1257
+ "qpos_pj": {
1258
+ "max": 2.8352200984954834,
1259
+ "mean": 0.1325104517579478,
1260
+ "min": 0.0033388733863830566,
1261
+ "std": 0.4389731704134287
1262
+ },
1263
+ "reference_body_jerk_level": 0.032168975994990016,
1264
+ "reference_qpos_jerk_level": 0.07551584170589375,
1265
+ "transition_window_length": 60,
1266
+ "transition_windows": [
1267
+ [
1268
+ 30,
1269
+ 90
1270
+ ]
1271
+ ]
1272
+ }
1273
+ },
1274
+ "omomo_original_filtered_test": {
1275
+ "embedding": {
1276
+ "diversity_generated": 1.1345490872611939,
1277
+ "motion_fid": 1.2325407801917518,
1278
+ "motion_kid": {
1279
+ "mean": 0.002298161219308703,
1280
+ "std": 0.0
1281
+ },
1282
+ "num_samples": 42
1283
+ },
1284
+ "num_samples": 42,
1285
+ "physical": {
1286
+ "body_jerk_mean": {
1287
+ "max": 215.9715118408203,
1288
+ "mean": 64.15830364681426,
1289
+ "min": 10.436027526855469,
1290
+ "std": 44.76208781970577
1291
+ },
1292
+ "contact_sliding_speed": {
1293
+ "max": 1.5735926628112793,
1294
+ "mean": 0.47663801983885823,
1295
+ "min": 0.015720246359705925,
1296
+ "std": 0.4178593234495149
1297
+ },
1298
+ "foot_ground_error": {
1299
+ "max": 0.01301527488976717,
1300
+ "mean": 0.006088023423217237,
1301
+ "min": 0.0020812139846384525,
1302
+ "std": 0.002377843957042252
1303
+ },
1304
+ "num_samples": 42
1305
+ },
1306
+ "text": {
1307
+ "generated": {
1308
+ "matching_distance": {
1309
+ "max": 1.433463454246521,
1310
+ "mean": 1.305754039968763,
1311
+ "min": 1.2346299886703491,
1312
+ "std": 0.042840413554061436
1313
+ },
1314
+ "matching_score": 1.3057539463043213,
1315
+ "mean_text_rank": 8.6875,
1316
+ "median_text_rank": 5.5,
1317
+ "r_precision": [
1318
+ 0.25,
1319
+ 0.3125,
1320
+ 0.40625
1321
+ ],
1322
+ "retrieval_batch_size": 32,
1323
+ "retrieval_num_batches": 1,
1324
+ "retrieval_num_samples": 32,
1325
+ "retrieval_order_policy": "manifest_order",
1326
+ "retrieval_unique_datasets_per_batch": {
1327
+ "max": 1.0,
1328
+ "mean": 1.0,
1329
+ "min": 1.0,
1330
+ "std": 0.0
1331
+ }
1332
+ },
1333
+ "reference": {
1334
+ "matching_distance": {
1335
+ "max": 1.3671505451202393,
1336
+ "mean": 1.269894642489297,
1337
+ "min": 1.2092769145965576,
1338
+ "std": 0.03030912014653846
1339
+ },
1340
+ "matching_score": 1.2698945999145508,
1341
+ "mean_text_rank": 8.40625,
1342
+ "median_text_rank": 6.5,
1343
+ "r_precision": [
1344
+ 0.15625,
1345
+ 0.21875,
1346
+ 0.25
1347
+ ],
1348
+ "retrieval_batch_size": 32,
1349
+ "retrieval_num_batches": 1,
1350
+ "retrieval_num_samples": 32,
1351
+ "retrieval_order_policy": "manifest_order",
1352
+ "retrieval_unique_datasets_per_batch": {
1353
+ "max": 1.0,
1354
+ "mean": 1.0,
1355
+ "min": 1.0,
1356
+ "std": 0.0
1357
+ }
1358
+ }
1359
+ },
1360
+ "transition": {
1361
+ "auj": {
1362
+ "max": 1.8620231315843276,
1363
+ "mean": 1.7725733755803437,
1364
+ "min": 1.6676089530947735,
1365
+ "std": 0.049378376046624706
1366
+ },
1367
+ "body_auj": {
1368
+ "max": 1.8620231315843276,
1369
+ "mean": 1.7725733755803437,
1370
+ "min": 1.6676089530947735,
1371
+ "std": 0.049378376046624706
1372
+ },
1373
+ "body_pj": {
1374
+ "max": 0.1809138059616089,
1375
+ "mean": 0.04543182488885664,
1376
+ "min": 0.004288120195269585,
1377
+ "std": 0.03565515372849116
1378
+ },
1379
+ "boundary_frames": [
1380
+ 60
1381
+ ],
1382
+ "chunk_length": 60,
1383
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1384
+ "num_frames": 120,
1385
+ "num_samples": 42,
1386
+ "pj": {
1387
+ "max": 0.1809138059616089,
1388
+ "mean": 0.04543182488885664,
1389
+ "min": 0.004288120195269585,
1390
+ "std": 0.03565515372849116
1391
+ },
1392
+ "qpos_auj": {
1393
+ "max": 28.471612076712674,
1394
+ "mean": 5.603556235023785,
1395
+ "min": 4.300355892869041,
1396
+ "std": 3.8283692334934805
1397
+ },
1398
+ "qpos_pj": {
1399
+ "max": 11.778656005859375,
1400
+ "mean": 0.6776692000927315,
1401
+ "min": 0.007280692458152771,
1402
+ "std": 1.8678563037401907
1403
+ },
1404
+ "reference_body_jerk_level": 0.032168975994990016,
1405
+ "reference_qpos_jerk_level": 0.07551584170589375,
1406
+ "transition_window_length": 60,
1407
+ "transition_windows": [
1408
+ [
1409
+ 30,
1410
+ 90
1411
+ ]
1412
+ ]
1413
+ }
1414
+ },
1415
+ "permo_original_filtered_test": {
1416
+ "embedding": {
1417
+ "diversity_generated": 1.0877195287239692,
1418
+ "motion_fid": 1.1566815401268762,
1419
+ "motion_kid": {
1420
+ "mean": 0.0022956662767517066,
1421
+ "std": 0.0
1422
+ },
1423
+ "num_samples": 42
1424
+ },
1425
+ "num_samples": 42,
1426
+ "physical": {
1427
+ "body_jerk_mean": {
1428
+ "max": 263.7925720214844,
1429
+ "mean": 76.4998429048629,
1430
+ "min": 4.8154168128967285,
1431
+ "std": 69.10691261564781
1432
+ },
1433
+ "contact_sliding_speed": {
1434
+ "max": 1.7778644561767578,
1435
+ "mean": 0.31765077484860305,
1436
+ "min": 0.009715426713228226,
1437
+ "std": 0.403836893708808
1438
+ },
1439
+ "foot_ground_error": {
1440
+ "max": 0.025011582300066948,
1441
+ "mean": 0.007566677107076559,
1442
+ "min": 0.0026211035437881947,
1443
+ "std": 0.004112016751390115
1444
+ },
1445
+ "num_samples": 42
1446
+ },
1447
+ "text": {
1448
+ "generated": {
1449
+ "matching_distance": {
1450
+ "max": 1.3902548551559448,
1451
+ "mean": 1.3313227948688326,
1452
+ "min": 1.2616204023361206,
1453
+ "std": 0.0340989370710566
1454
+ },
1455
+ "matching_score": 1.3313229084014893,
1456
+ "mean_text_rank": 4.84375,
1457
+ "median_text_rank": 4.0,
1458
+ "r_precision": [
1459
+ 0.375,
1460
+ 0.40625,
1461
+ 0.4375
1462
+ ],
1463
+ "retrieval_batch_size": 32,
1464
+ "retrieval_num_batches": 1,
1465
+ "retrieval_num_samples": 32,
1466
+ "retrieval_order_policy": "manifest_order",
1467
+ "retrieval_unique_datasets_per_batch": {
1468
+ "max": 1.0,
1469
+ "mean": 1.0,
1470
+ "min": 1.0,
1471
+ "std": 0.0
1472
+ }
1473
+ },
1474
+ "reference": {
1475
+ "matching_distance": {
1476
+ "max": 1.3314403295516968,
1477
+ "mean": 1.2064374969119118,
1478
+ "min": 1.1483595371246338,
1479
+ "std": 0.03169944929872626
1480
+ },
1481
+ "matching_score": 1.2064374685287476,
1482
+ "mean_text_rank": 4.34375,
1483
+ "median_text_rank": 3.0,
1484
+ "r_precision": [
1485
+ 0.25,
1486
+ 0.4375,
1487
+ 0.53125
1488
+ ],
1489
+ "retrieval_batch_size": 32,
1490
+ "retrieval_num_batches": 1,
1491
+ "retrieval_num_samples": 32,
1492
+ "retrieval_order_policy": "manifest_order",
1493
+ "retrieval_unique_datasets_per_batch": {
1494
+ "max": 1.0,
1495
+ "mean": 1.0,
1496
+ "min": 1.0,
1497
+ "std": 0.0
1498
+ }
1499
+ }
1500
+ },
1501
+ "transition": {
1502
+ "auj": {
1503
+ "max": 2.781857592675066,
1504
+ "mean": 1.8609303950500373,
1505
+ "min": 1.5932010662360152,
1506
+ "std": 0.2132839801381445
1507
+ },
1508
+ "body_auj": {
1509
+ "max": 2.781857592675066,
1510
+ "mean": 1.8609303950500373,
1511
+ "min": 1.5932010662360152,
1512
+ "std": 0.2132839801381445
1513
+ },
1514
+ "body_pj": {
1515
+ "max": 0.2385309636592865,
1516
+ "mean": 0.0728943720903425,
1517
+ "min": 0.0014599980786442757,
1518
+ "std": 0.0661067180136749
1519
+ },
1520
+ "boundary_frames": [
1521
+ 60
1522
+ ],
1523
+ "chunk_length": 60,
1524
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1525
+ "num_frames": 120,
1526
+ "num_samples": 42,
1527
+ "pj": {
1528
+ "max": 0.2385309636592865,
1529
+ "mean": 0.0728943720903425,
1530
+ "min": 0.0014599980786442757,
1531
+ "std": 0.0661067180136749
1532
+ },
1533
+ "qpos_auj": {
1534
+ "max": 12.99549435606882,
1535
+ "mean": 5.456390068404779,
1536
+ "min": 4.300785458770909,
1537
+ "std": 1.9640080094470993
1538
+ },
1539
+ "qpos_pj": {
1540
+ "max": 3.9514918252825737,
1541
+ "mean": 0.4679690366056526,
1542
+ "min": 0.006184376776218414,
1543
+ "std": 0.7646300858697805
1544
+ },
1545
+ "reference_body_jerk_level": 0.032168975994990016,
1546
+ "reference_qpos_jerk_level": 0.07551584170589375,
1547
+ "transition_window_length": 60,
1548
+ "transition_windows": [
1549
+ [
1550
+ 30,
1551
+ 90
1552
+ ]
1553
+ ]
1554
+ }
1555
+ },
1556
+ "snapmogen_original_filtered_test": {
1557
+ "embedding": {
1558
+ "diversity_generated": 1.2069637839684089,
1559
+ "motion_fid": 1.3497101089407968,
1560
+ "motion_kid": {
1561
+ "mean": 0.0020190263621038262,
1562
+ "std": 0.0
1563
+ },
1564
+ "num_samples": 42
1565
+ },
1566
+ "num_samples": 42,
1567
+ "physical": {
1568
+ "body_jerk_mean": {
1569
+ "max": 228.76815795898438,
1570
+ "mean": 39.739420181229,
1571
+ "min": 2.503999710083008,
1572
+ "std": 55.302429574414276
1573
+ },
1574
+ "contact_sliding_speed": {
1575
+ "max": 1.628731369972229,
1576
+ "mean": 0.2116111345655684,
1577
+ "min": 0.006765882950276136,
1578
+ "std": 0.36438021456289116
1579
+ },
1580
+ "foot_ground_error": {
1581
+ "max": 0.011111666448414326,
1582
+ "mean": 0.005569282187415021,
1583
+ "min": 0.0005915897199884057,
1584
+ "std": 0.0024525098901194337
1585
+ },
1586
+ "num_samples": 42
1587
+ },
1588
+ "text": {
1589
+ "generated": {
1590
+ "matching_distance": {
1591
+ "max": 1.4769279956817627,
1592
+ "mean": 1.377476453781128,
1593
+ "min": 1.2130833864212036,
1594
+ "std": 0.046379623782035695
1595
+ },
1596
+ "matching_score": 1.377476453781128,
1597
+ "mean_text_rank": 6.84375,
1598
+ "median_text_rank": 4.0,
1599
+ "r_precision": [
1600
+ 0.34375,
1601
+ 0.4375,
1602
+ 0.46875
1603
+ ],
1604
+ "retrieval_batch_size": 32,
1605
+ "retrieval_num_batches": 1,
1606
+ "retrieval_num_samples": 32,
1607
+ "retrieval_order_policy": "manifest_order",
1608
+ "retrieval_unique_datasets_per_batch": {
1609
+ "max": 1.0,
1610
+ "mean": 1.0,
1611
+ "min": 1.0,
1612
+ "std": 0.0
1613
+ }
1614
+ },
1615
+ "reference": {
1616
+ "matching_distance": {
1617
+ "max": 1.2890117168426514,
1618
+ "mean": 1.2102941047577631,
1619
+ "min": 1.1285470724105835,
1620
+ "std": 0.04065653924035539
1621
+ },
1622
+ "matching_score": 1.2102941274642944,
1623
+ "mean_text_rank": 1.75,
1624
+ "median_text_rank": 1.0,
1625
+ "r_precision": [
1626
+ 0.75,
1627
+ 0.8125,
1628
+ 0.84375
1629
+ ],
1630
+ "retrieval_batch_size": 32,
1631
+ "retrieval_num_batches": 1,
1632
+ "retrieval_num_samples": 32,
1633
+ "retrieval_order_policy": "manifest_order",
1634
+ "retrieval_unique_datasets_per_batch": {
1635
+ "max": 1.0,
1636
+ "mean": 1.0,
1637
+ "min": 1.0,
1638
+ "std": 0.0
1639
+ }
1640
+ }
1641
+ },
1642
+ "transition": {
1643
+ "auj": {
1644
+ "max": 5.563047105616452,
1645
+ "mean": 1.9004562439177437,
1646
+ "min": 1.720805086064156,
1647
+ "std": 0.5739356354454959
1648
+ },
1649
+ "body_auj": {
1650
+ "max": 5.563047105616452,
1651
+ "mean": 1.9004562439177437,
1652
+ "min": 1.720805086064156,
1653
+ "std": 0.5739356354454959
1654
+ },
1655
+ "body_pj": {
1656
+ "max": 1.3313361704349518,
1657
+ "mean": 0.058183968610440694,
1658
+ "min": 0.0004454776644706726,
1659
+ "std": 0.20293248270696373
1660
+ },
1661
+ "boundary_frames": [
1662
+ 60
1663
+ ],
1664
+ "chunk_length": 60,
1665
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1666
+ "num_frames": 120,
1667
+ "num_samples": 42,
1668
+ "pj": {
1669
+ "max": 1.3313361704349518,
1670
+ "mean": 0.058183968610440694,
1671
+ "min": 0.0004454776644706726,
1672
+ "std": 0.20293248270696373
1673
+ },
1674
+ "qpos_auj": {
1675
+ "max": 52.49802010814777,
1676
+ "mean": 6.607324879388726,
1677
+ "min": 4.299905554166392,
1678
+ "std": 8.780425573903713
1679
+ },
1680
+ "qpos_pj": {
1681
+ "max": 12.550551891326904,
1682
+ "mean": 0.81523050702942,
1683
+ "min": 0.0026228129863739014,
1684
+ "std": 2.5778904370957694
1685
+ },
1686
+ "reference_body_jerk_level": 0.032168975994990016,
1687
+ "reference_qpos_jerk_level": 0.07551584170589375,
1688
+ "transition_window_length": 60,
1689
+ "transition_windows": [
1690
+ [
1691
+ 30,
1692
+ 90
1693
+ ]
1694
+ ]
1695
+ }
1696
+ }
1697
+ },
1698
+ "embedding_frames": 120,
1699
+ "embedding_metrics": {
1700
+ "diversity_generated": 1.185870855502696,
1701
+ "embedding_frames": 120,
1702
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_040000.pt",
1703
+ "matching_score_generated": 1.3311593532562256,
1704
+ "matching_score_reference": 1.244696855545044,
1705
+ "motion_fid": 0.5886926257750928,
1706
+ "motion_key": "body_pos_local",
1707
+ "motion_kid": {
1708
+ "mean": 0.0009367047902468784,
1709
+ "std": 0.0
1710
+ },
1711
+ "multimodality": null,
1712
+ "num_samples": 512,
1713
+ "r_precision_generated": [
1714
+ 0.240234375,
1715
+ 0.33984375,
1716
+ 0.412109375
1717
+ ],
1718
+ "r_precision_reference": [
1719
+ 0.705078125,
1720
+ 0.8359375,
1721
+ 0.90234375
1722
+ ],
1723
+ "text_cache_dir": "outputs/evaluator_text_cache/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f",
1724
+ "text_encoder_model": "${OMG_MODELS_ROOT}/text_encoders/t5-3b-local",
1725
+ "text_generated": {
1726
+ "matching_distance": {
1727
+ "max": 1.4769279956817627,
1728
+ "mean": 1.331159382360056,
1729
+ "min": 1.134010910987854,
1730
+ "std": 0.06046604617699041
1731
+ },
1732
+ "matching_score": 1.3311593532562256,
1733
+ "mean_text_rank": 8.705078125,
1734
+ "median_text_rank": 6.0,
1735
+ "r_precision": [
1736
+ 0.240234375,
1737
+ 0.33984375,
1738
+ 0.412109375
1739
+ ],
1740
+ "retrieval_batch_size": 32,
1741
+ "retrieval_num_batches": 16,
1742
+ "retrieval_num_samples": 512,
1743
+ "retrieval_order_policy": "stratified_by_dataset",
1744
+ "retrieval_unique_datasets_per_batch": {
1745
+ "max": 12.0,
1746
+ "mean": 12.0,
1747
+ "min": 12.0,
1748
+ "std": 0.0
1749
+ }
1750
+ },
1751
+ "text_reference": {
1752
+ "matching_distance": {
1753
+ "max": 1.4135501384735107,
1754
+ "mean": 1.2446968206204474,
1755
+ "min": 1.1285470724105835,
1756
+ "std": 0.0457579048836768
1757
+ },
1758
+ "matching_score": 1.244696855545044,
1759
+ "mean_text_rank": 1.869140625,
1760
+ "median_text_rank": 1.0,
1761
+ "r_precision": [
1762
+ 0.705078125,
1763
+ 0.8359375,
1764
+ 0.90234375
1765
+ ],
1766
+ "retrieval_batch_size": 32,
1767
+ "retrieval_num_batches": 16,
1768
+ "retrieval_num_samples": 512,
1769
+ "retrieval_order_policy": "stratified_by_dataset",
1770
+ "retrieval_unique_datasets_per_batch": {
1771
+ "max": 12.0,
1772
+ "mean": 12.0,
1773
+ "min": 12.0,
1774
+ "std": 0.0
1775
+ }
1776
+ },
1777
+ "text_retrieval_batch_size": 32
1778
+ },
1779
+ "exp": "g1_transformer_guided_300m",
1780
+ "generated_qpos": "${OMG_REPO_ROOT}/outputs_baseline/kimodo/g1_seed/text_mixed_modalities_all_v1_120f_transition/generated_qpos.npz",
1781
+ "label": "Kimodo-G1-SEED-v1",
1782
+ "num_frames": 120,
1783
+ "num_samples": 512,
1784
+ "physical_metrics": {
1785
+ "body_jerk_mean": {
1786
+ "max": 531.6823120117188,
1787
+ "mean": 72.15493237087503,
1788
+ "min": 2.1355345249176025,
1789
+ "std": 79.05384587679872
1790
+ },
1791
+ "contact_height_threshold": 0.12,
1792
+ "contact_penetration_tolerance": 0.02,
1793
+ "contact_sliding_speed": {
1794
+ "max": 2.914012908935547,
1795
+ "mean": 0.3516153497353116,
1796
+ "min": 0.0,
1797
+ "std": 0.475502377479124
1798
+ },
1799
+ "foot_ground_error": {
1800
+ "max": 1.477441668510437,
1801
+ "mean": 0.009914551770975777,
1802
+ "min": 0.0005915897199884057,
1803
+ "std": 0.06528683401820799
1804
+ },
1805
+ "frames": 120,
1806
+ "num_samples": 512
1807
+ },
1808
+ "real_motion_metrics": {
1809
+ "diversity_real": 1.322922111091206,
1810
+ "matching_distance_real": {
1811
+ "max": 1.4135501384735107,
1812
+ "mean": 1.2446968206204474,
1813
+ "min": 1.1285470724105835,
1814
+ "std": 0.0457579048836768
1815
+ },
1816
+ "matching_score_real": 1.244696855545044,
1817
+ "motion_fid_reference_eval": 0.0,
1818
+ "motion_fid_split_half": 0.0,
1819
+ "motion_kid_reference_eval": {
1820
+ "mean": -1.9876635613869098e-05,
1821
+ "std": 0.0
1822
+ },
1823
+ "motion_kid_split_half": {
1824
+ "mean": -1.9876635613869098e-05,
1825
+ "std": 0.0
1826
+ },
1827
+ "multimodality": "-",
1828
+ "num_reference_samples": 512,
1829
+ "num_samples": 512,
1830
+ "r_precision_real": [
1831
+ 0.455078125,
1832
+ 0.587890625,
1833
+ 0.677734375
1834
+ ],
1835
+ "text_real": {
1836
+ "matching_distance": {
1837
+ "max": 1.4135501384735107,
1838
+ "mean": 1.2446968206204474,
1839
+ "min": 1.1285470724105835,
1840
+ "std": 0.0457579048836768
1841
+ },
1842
+ "matching_score": 1.244696855545044,
1843
+ "mean_text_rank": 3.904296875,
1844
+ "median_text_rank": 2.0,
1845
+ "r_precision": [
1846
+ 0.455078125,
1847
+ 0.587890625,
1848
+ 0.677734375
1849
+ ],
1850
+ "retrieval_batch_size": 32,
1851
+ "retrieval_num_batches": 16,
1852
+ "retrieval_num_samples": 512,
1853
+ "retrieval_order_policy": "manifest_order",
1854
+ "retrieval_unique_datasets_per_batch": null
1855
+ }
1856
+ },
1857
+ "reference_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/reference_qpos.npz",
1858
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/sample_metrics.jsonl",
1859
+ "sample_rankings_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/sample_rankings.json",
1860
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_samples/mixed_modalities_all_v1/text_test_512_120f.jsonl",
1861
+ "split": "test",
1862
+ "tracker_executed": {
1863
+ "enabled": false
1864
+ },
1865
+ "transition_metrics": {
1866
+ "auj": {
1867
+ "max": 6.514222457101631,
1868
+ "mean": 1.8406120086684412,
1869
+ "min": 1.4860971376024386,
1870
+ "std": 0.3403637125100069
1871
+ },
1872
+ "body_auj": {
1873
+ "max": 6.514222457101631,
1874
+ "mean": 1.8406120086684412,
1875
+ "min": 1.4860971376024386,
1876
+ "std": 0.3403637125100069
1877
+ },
1878
+ "body_pj": {
1879
+ "max": 1.9185047969222069,
1880
+ "mean": 0.0578470742943864,
1881
+ "min": 0.0004047006368637085,
1882
+ "std": 0.12257751816949211
1883
+ },
1884
+ "boundary_frames": [
1885
+ 60
1886
+ ],
1887
+ "chunk_length": 60,
1888
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
1889
+ "enabled": true,
1890
+ "num_frames": 120,
1891
+ "num_samples": 512,
1892
+ "pj": {
1893
+ "max": 1.9185047969222069,
1894
+ "mean": 0.0578470742943864,
1895
+ "min": 0.0004047006368637085,
1896
+ "std": 0.12257751816949211
1897
+ },
1898
+ "qpos_auj": {
1899
+ "max": 144.49321365365512,
1900
+ "mean": 6.459746310243598,
1901
+ "min": 4.2940439462623115,
1902
+ "std": 8.748215194373627
1903
+ },
1904
+ "qpos_pj": {
1905
+ "max": 12.888956546783447,
1906
+ "mean": 0.706420054541411,
1907
+ "min": 0.0026228129863739014,
1908
+ "std": 1.8797036959973208
1909
+ },
1910
+ "reference_body_jerk_level": 0.032168975994990016,
1911
+ "reference_qpos_jerk_level": 0.07551584170589375,
1912
+ "transition_window_length": 60,
1913
+ "transition_windows": [
1914
+ [
1915
+ 30,
1916
+ 90
1917
+ ]
1918
+ ]
1919
+ }
1920
+ },
1921
+ "label": "Kimodo-G1-SEED-v1"
1922
+ }
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/physical_metrics.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "body_jerk_mean": {
3
+ "max": 531.6823120117188,
4
+ "mean": 72.15493237087503,
5
+ "min": 2.1355345249176025,
6
+ "std": 79.05384587679872
7
+ },
8
+ "contact_height_threshold": 0.12,
9
+ "contact_penetration_tolerance": 0.02,
10
+ "contact_sliding_speed": {
11
+ "max": 2.914012908935547,
12
+ "mean": 0.3516153497353116,
13
+ "min": 0.0,
14
+ "std": 0.475502377479124
15
+ },
16
+ "foot_ground_error": {
17
+ "max": 1.477441668510437,
18
+ "mean": 0.009914551770975777,
19
+ "min": 0.0005915897199884057,
20
+ "std": 0.06528683401820799
21
+ },
22
+ "frames": 120,
23
+ "num_samples": 512
24
+ }
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/real_motion_metrics.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "diversity_real": 1.322922111091206,
3
+ "matching_distance_real": {
4
+ "max": 1.4135501384735107,
5
+ "mean": 1.2446968206204474,
6
+ "min": 1.1285470724105835,
7
+ "std": 0.0457579048836768
8
+ },
9
+ "matching_score_real": 1.244696855545044,
10
+ "motion_fid_reference_eval": 0.0,
11
+ "motion_fid_split_half": 0.0,
12
+ "motion_kid_reference_eval": {
13
+ "mean": -1.9876635613869098e-05,
14
+ "std": 0.0
15
+ },
16
+ "motion_kid_split_half": {
17
+ "mean": -1.9876635613869098e-05,
18
+ "std": 0.0
19
+ },
20
+ "multimodality": "-",
21
+ "num_reference_samples": 512,
22
+ "num_samples": 512,
23
+ "r_precision_real": [
24
+ 0.455078125,
25
+ 0.587890625,
26
+ 0.677734375
27
+ ],
28
+ "text_real": {
29
+ "matching_distance": {
30
+ "max": 1.4135501384735107,
31
+ "mean": 1.2446968206204474,
32
+ "min": 1.1285470724105835,
33
+ "std": 0.0457579048836768
34
+ },
35
+ "matching_score": 1.244696855545044,
36
+ "mean_text_rank": 3.904296875,
37
+ "median_text_rank": 2.0,
38
+ "r_precision": [
39
+ 0.455078125,
40
+ 0.587890625,
41
+ 0.677734375
42
+ ],
43
+ "retrieval_batch_size": 32,
44
+ "retrieval_num_batches": 16,
45
+ "retrieval_num_samples": 512,
46
+ "retrieval_order_policy": "manifest_order",
47
+ "retrieval_unique_datasets_per_batch": null
48
+ }
49
+ }
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/sample_metrics.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/sample_rankings.json ADDED
@@ -0,0 +1,1113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_generated": [
3
+ {
4
+ "caption": "character points back over left shoulder then starts walking forward and points on the ground",
5
+ "dataset": "bones_seed_original_filtered_test",
6
+ "dataset_index": 16469,
7
+ "generated_matching_distance": 1.134010910987854,
8
+ "generated_reference_distance_gap": -0.033188819885253906,
9
+ "generated_text_rank": 1,
10
+ "reference_matching_distance": 1.167199730873108,
11
+ "reference_text_rank": 1,
12
+ "sample_index": 111
13
+ },
14
+ {
15
+ "caption": "character sorts out medium heavy object with one hand from low behind and puts it away to the behind high",
16
+ "dataset": "bones_seed_original_filtered_test",
17
+ "dataset_index": 14825,
18
+ "generated_matching_distance": 1.1501623392105103,
19
+ "generated_reference_distance_gap": -0.004802584648132324,
20
+ "generated_text_rank": 1,
21
+ "reference_matching_distance": 1.1549649238586426,
22
+ "reference_text_rank": 1,
23
+ "sample_index": 107
24
+ },
25
+ {
26
+ "caption": "person is executing a swift turn to the right in place, beginning with their right foot",
27
+ "dataset": "bones_seed_original_filtered_test",
28
+ "dataset_index": 9020,
29
+ "generated_matching_distance": 1.173611044883728,
30
+ "generated_reference_distance_gap": -0.01781463623046875,
31
+ "generated_text_rank": 1,
32
+ "reference_matching_distance": 1.1914256811141968,
33
+ "reference_text_rank": 1,
34
+ "sample_index": 102
35
+ },
36
+ {
37
+ "caption": "walking forward with hands placed behind the back",
38
+ "dataset": "bones_seed_original_filtered_test",
39
+ "dataset_index": 21257,
40
+ "generated_matching_distance": 1.1791802644729614,
41
+ "generated_reference_distance_gap": -0.0008120536804199219,
42
+ "generated_text_rank": 1,
43
+ "reference_matching_distance": 1.1799923181533813,
44
+ "reference_text_rank": 1,
45
+ "sample_index": 124
46
+ },
47
+ {
48
+ "caption": "start jogging forward",
49
+ "dataset": "bones_seed_original_filtered_test",
50
+ "dataset_index": 11651,
51
+ "generated_matching_distance": 1.1811357736587524,
52
+ "generated_reference_distance_gap": -0.009389281272888184,
53
+ "generated_text_rank": 1,
54
+ "reference_matching_distance": 1.1905250549316406,
55
+ "reference_text_rank": 1,
56
+ "sample_index": 105
57
+ },
58
+ {
59
+ "caption": "person stops and walks backwards",
60
+ "dataset": "bones_seed_original_filtered_test",
61
+ "dataset_index": 21040,
62
+ "generated_matching_distance": 1.1818833351135254,
63
+ "generated_reference_distance_gap": -0.00507962703704834,
64
+ "generated_text_rank": 1,
65
+ "reference_matching_distance": 1.1869629621505737,
66
+ "reference_text_rank": 1,
67
+ "sample_index": 123
68
+ },
69
+ {
70
+ "caption": "running forward in one place dance move",
71
+ "dataset": "bones_seed_original_filtered_test",
72
+ "dataset_index": 4937,
73
+ "generated_matching_distance": 1.1956936120986938,
74
+ "generated_reference_distance_gap": -0.08252894878387451,
75
+ "generated_text_rank": 1,
76
+ "reference_matching_distance": 1.2782225608825684,
77
+ "reference_text_rank": 1,
78
+ "sample_index": 95
79
+ },
80
+ {
81
+ "caption": "individual is depicted painting on a canvas",
82
+ "dataset": "bones_seed_original_filtered_test",
83
+ "dataset_index": 16065,
84
+ "generated_matching_distance": 1.1977249383926392,
85
+ "generated_reference_distance_gap": -0.029548168182373047,
86
+ "generated_text_rank": 1,
87
+ "reference_matching_distance": 1.2272731065750122,
88
+ "reference_text_rank": 1,
89
+ "sample_index": 109
90
+ },
91
+ {
92
+ "caption": "open a fridge, take out an item from a low shelf, and close the fridge using the right hand",
93
+ "dataset": "bones_seed_original_filtered_test",
94
+ "dataset_index": 15704,
95
+ "generated_matching_distance": 1.1987175941467285,
96
+ "generated_reference_distance_gap": 0.029857873916625977,
97
+ "generated_text_rank": 1,
98
+ "reference_matching_distance": 1.1688597202301025,
99
+ "reference_text_rank": 1,
100
+ "sample_index": 108
101
+ },
102
+ {
103
+ "caption": "crouching and walking to the right in a stealthy manner",
104
+ "dataset": "bones_seed_original_filtered_test",
105
+ "dataset_index": 2787,
106
+ "generated_matching_distance": 1.2028462886810303,
107
+ "generated_reference_distance_gap": -0.024004817008972168,
108
+ "generated_text_rank": 1,
109
+ "reference_matching_distance": 1.2268511056900024,
110
+ "reference_text_rank": 1,
111
+ "sample_index": 90
112
+ },
113
+ {
114
+ "caption": "person quickly performs a vogue dance catwalk move facing forward",
115
+ "dataset": "bones_seed_original_filtered_test",
116
+ "dataset_index": 4881,
117
+ "generated_matching_distance": 1.2047133445739746,
118
+ "generated_reference_distance_gap": 0.03519153594970703,
119
+ "generated_text_rank": 1,
120
+ "reference_matching_distance": 1.1695218086242676,
121
+ "reference_text_rank": 1,
122
+ "sample_index": 94
123
+ },
124
+ {
125
+ "caption": "character sorts out medium big heavy object with one hand from high up front and puts it away up front low height",
126
+ "dataset": "bones_seed_original_filtered_test",
127
+ "dataset_index": 14620,
128
+ "generated_matching_distance": 1.2063437700271606,
129
+ "generated_reference_distance_gap": -0.005927562713623047,
130
+ "generated_text_rank": 1,
131
+ "reference_matching_distance": 1.2122713327407837,
132
+ "reference_text_rank": 1,
133
+ "sample_index": 106
134
+ },
135
+ {
136
+ "caption": "person performs a quick hip hop padebure by switching feet at a tempo of 120 bpm",
137
+ "dataset": "bones_seed_original_filtered_test",
138
+ "dataset_index": 3935,
139
+ "generated_matching_distance": 1.2069758176803589,
140
+ "generated_reference_distance_gap": 0.001990079879760742,
141
+ "generated_text_rank": 1,
142
+ "reference_matching_distance": 1.2049857378005981,
143
+ "reference_text_rank": 1,
144
+ "sample_index": 93
145
+ },
146
+ {
147
+ "caption": "person walking backwards diagonally to the left",
148
+ "dataset": "bones_seed_original_filtered_test",
149
+ "dataset_index": 21347,
150
+ "generated_matching_distance": 1.2078200578689575,
151
+ "generated_reference_distance_gap": -0.02900385856628418,
152
+ "generated_text_rank": 1,
153
+ "reference_matching_distance": 1.2368239164352417,
154
+ "reference_text_rank": 1,
155
+ "sample_index": 126
156
+ },
157
+ {
158
+ "caption": "start walking backwards left",
159
+ "dataset": "bones_seed_original_filtered_test",
160
+ "dataset_index": 21369,
161
+ "generated_matching_distance": 1.209242343902588,
162
+ "generated_reference_distance_gap": 0.010047316551208496,
163
+ "generated_text_rank": 1,
164
+ "reference_matching_distance": 1.1991950273513794,
165
+ "reference_text_rank": 1,
166
+ "sample_index": 127
167
+ },
168
+ {
169
+ "caption": "person ceases walking forward, placing hands behind their back with right hand closer to body",
170
+ "dataset": "bones_seed_original_filtered_test",
171
+ "dataset_index": 21302,
172
+ "generated_matching_distance": 1.2130671739578247,
173
+ "generated_reference_distance_gap": -0.020138859748840332,
174
+ "generated_text_rank": 1,
175
+ "reference_matching_distance": 1.233206033706665,
176
+ "reference_text_rank": 1,
177
+ "sample_index": 125
178
+ },
179
+ {
180
+ "caption": "A person is grabbing something from a shelf. They reach upward on the toes with their left arm raised above their head, and their head tilted up. They make one step forward with their left leg. Their right arm is slightly aside.",
181
+ "dataset": "snapmogen_original_filtered_test",
182
+ "dataset_index": 8554,
183
+ "generated_matching_distance": 1.2130833864212036,
184
+ "generated_reference_distance_gap": 0.03702831268310547,
185
+ "generated_text_rank": 1,
186
+ "reference_matching_distance": 1.1760550737380981,
187
+ "reference_text_rank": 1,
188
+ "sample_index": 500
189
+ },
190
+ {
191
+ "caption": "a jog in a clockwise circle",
192
+ "dataset": "bones_seed_original_filtered_test",
193
+ "dataset_index": 11006,
194
+ "generated_matching_distance": 1.216279149055481,
195
+ "generated_reference_distance_gap": 0.0027948617935180664,
196
+ "generated_text_rank": 1,
197
+ "reference_matching_distance": 1.213484287261963,
198
+ "reference_text_rank": 1,
199
+ "sample_index": 104
200
+ },
201
+ {
202
+ "caption": "run continuously with turns, spins, and direction shifts; style: dynamic, agile",
203
+ "dataset": "lafan1_original_filtered_test",
204
+ "dataset_index": 152,
205
+ "generated_matching_distance": 1.22147798538208,
206
+ "generated_reference_distance_gap": 0.0435643196105957,
207
+ "generated_text_rank": 1,
208
+ "reference_matching_distance": 1.1779136657714844,
209
+ "reference_text_rank": 1,
210
+ "sample_index": 275
211
+ },
212
+ {
213
+ "caption": "character expresses ignorance by shrugging without using their hands",
214
+ "dataset": "bones_seed_original_filtered_test",
215
+ "dataset_index": 6879,
216
+ "generated_matching_distance": 1.2220014333724976,
217
+ "generated_reference_distance_gap": -0.0694953203201294,
218
+ "generated_text_rank": 1,
219
+ "reference_matching_distance": 1.291496753692627,
220
+ "reference_text_rank": 2,
221
+ "sample_index": 98
222
+ }
223
+ ],
224
+ "best_reference": [
225
+ {
226
+ "caption": "The person walks forward in a depressed and sluggish manner, tilting their head to the left and leaning their torso slightly forward, moving their arms gently, and taking short, slow steps.",
227
+ "dataset": "snapmogen_original_filtered_test",
228
+ "dataset_index": 9460,
229
+ "generated_matching_distance": 1.3915009498596191,
230
+ "generated_reference_distance_gap": 0.26295387744903564,
231
+ "generated_text_rank": 18,
232
+ "reference_matching_distance": 1.1285470724105835,
233
+ "reference_text_rank": 1,
234
+ "sample_index": 506
235
+ },
236
+ {
237
+ "caption": "The person walks in a circle. He walks in a circle to the right, turning slightly over his right shoulder, while waving his arms back and forth in a relaxed manner.",
238
+ "dataset": "snapmogen_original_filtered_test",
239
+ "dataset_index": 2114,
240
+ "generated_matching_distance": 1.385533094406128,
241
+ "generated_reference_distance_gap": 0.24202275276184082,
242
+ "generated_text_rank": 14,
243
+ "reference_matching_distance": 1.143510341644287,
244
+ "reference_text_rank": 1,
245
+ "sample_index": 476
246
+ },
247
+ {
248
+ "caption": "The person rises from her knees to her feet. The person kneels with her arms spread wide. She touches her palms together and tilts her head down. Then she rests her hands on the floor, leans on her left leg, and straightens up.",
249
+ "dataset": "snapmogen_original_filtered_test",
250
+ "dataset_index": 7444,
251
+ "generated_matching_distance": 1.320056676864624,
252
+ "generated_reference_distance_gap": 0.17481625080108643,
253
+ "generated_text_rank": 4,
254
+ "reference_matching_distance": 1.1452404260635376,
255
+ "reference_text_rank": 1,
256
+ "sample_index": 496
257
+ },
258
+ {
259
+ "caption": "The person is practicing low kicks with their right leg. They lift it up with the knee bent, raising their bend arms in front of them and leaning to the left a bit for balance. They kick forward and a bit to the right with their right foot repeatedly.",
260
+ "dataset": "snapmogen_original_filtered_test",
261
+ "dataset_index": 7248,
262
+ "generated_matching_distance": 1.358250617980957,
263
+ "generated_reference_distance_gap": 0.2123507261276245,
264
+ "generated_text_rank": 14,
265
+ "reference_matching_distance": 1.1458998918533325,
266
+ "reference_text_rank": 1,
267
+ "sample_index": 495
268
+ },
269
+ {
270
+ "caption": "sks person walks forward, then shifts sideways to alter their path.",
271
+ "dataset": "permo_original_filtered_test",
272
+ "dataset_index": 878,
273
+ "generated_matching_distance": 1.3643834590911865,
274
+ "generated_reference_distance_gap": 0.21602392196655273,
275
+ "generated_text_rank": 15,
276
+ "reference_matching_distance": 1.1483595371246338,
277
+ "reference_text_rank": 1,
278
+ "sample_index": 458
279
+ },
280
+ {
281
+ "caption": "a person jumps forward once.",
282
+ "dataset": "humanml_original_filtered_test",
283
+ "dataset_index": 5979,
284
+ "generated_matching_distance": 1.2455915212631226,
285
+ "generated_reference_distance_gap": 0.09371137619018555,
286
+ "generated_text_rank": 1,
287
+ "reference_matching_distance": 1.151880145072937,
288
+ "reference_text_rank": 1,
289
+ "sample_index": 204
290
+ },
291
+ {
292
+ "caption": "The person stands and raises their arms to head level and lowers them, then raises their left arm above their head and lowers it, and repeats this action.",
293
+ "dataset": "snapmogen_original_filtered_test",
294
+ "dataset_index": 2575,
295
+ "generated_matching_distance": 1.3383948802947998,
296
+ "generated_reference_distance_gap": 0.18643033504486084,
297
+ "generated_text_rank": 6,
298
+ "reference_matching_distance": 1.151964545249939,
299
+ "reference_text_rank": 1,
300
+ "sample_index": 479
301
+ },
302
+ {
303
+ "caption": "A person is doing overhead trap raises on the ground.",
304
+ "dataset": "motionllama_original_filtered_test",
305
+ "dataset_index": 308,
306
+ "generated_matching_distance": 1.3227300643920898,
307
+ "generated_reference_distance_gap": 0.16989505290985107,
308
+ "generated_text_rank": 2,
309
+ "reference_matching_distance": 1.1528350114822388,
310
+ "reference_text_rank": 1,
311
+ "sample_index": 354
312
+ },
313
+ {
314
+ "caption": "character sorts out medium heavy object with one hand from low behind and puts it away to the behind high",
315
+ "dataset": "bones_seed_original_filtered_test",
316
+ "dataset_index": 14825,
317
+ "generated_matching_distance": 1.1501623392105103,
318
+ "generated_reference_distance_gap": -0.004802584648132324,
319
+ "generated_text_rank": 1,
320
+ "reference_matching_distance": 1.1549649238586426,
321
+ "reference_text_rank": 1,
322
+ "sample_index": 107
323
+ },
324
+ {
325
+ "caption": "They are doing squats.",
326
+ "dataset": "motionllama_original_filtered_test",
327
+ "dataset_index": 92,
328
+ "generated_matching_distance": 1.3509737253189087,
329
+ "generated_reference_distance_gap": 0.19418644905090332,
330
+ "generated_text_rank": 2,
331
+ "reference_matching_distance": 1.1567872762680054,
332
+ "reference_text_rank": 1,
333
+ "sample_index": 346
334
+ },
335
+ {
336
+ "caption": "A person is pushing a chair with both hands.",
337
+ "dataset": "motionllama_original_filtered_test",
338
+ "dataset_index": 908,
339
+ "generated_matching_distance": 1.3709038496017456,
340
+ "generated_reference_distance_gap": 0.21327877044677734,
341
+ "generated_text_rank": 6,
342
+ "reference_matching_distance": 1.1576250791549683,
343
+ "reference_text_rank": 1,
344
+ "sample_index": 372
345
+ },
346
+ {
347
+ "caption": "A person stands with arms at its sides, lifts its arms laterally to horizontal position, then lowers them back to the starting position in a smooth, repetitive motion.; style: rhythmic smooth",
348
+ "dataset": "idea400_original_filtered_test",
349
+ "dataset_index": 342,
350
+ "generated_matching_distance": 1.366390347480774,
351
+ "generated_reference_distance_gap": 0.20729148387908936,
352
+ "generated_text_rank": 17,
353
+ "reference_matching_distance": 1.1590988636016846,
354
+ "reference_text_rank": 1,
355
+ "sample_index": 221
356
+ },
357
+ {
358
+ "caption": "The person jumps with their legs wide apart and arms actively raised above their head. Then the person takes a neutral resting position.",
359
+ "dataset": "snapmogen_original_filtered_test",
360
+ "dataset_index": 11567,
361
+ "generated_matching_distance": 1.3543835878372192,
362
+ "generated_reference_distance_gap": 0.1950664520263672,
363
+ "generated_text_rank": 7,
364
+ "reference_matching_distance": 1.159317135810852,
365
+ "reference_text_rank": 1,
366
+ "sample_index": 510
367
+ },
368
+ {
369
+ "caption": "The person stands with his head turned to the left and his left arm raised and bent at the elbow, waves his palm in front of him from left to right, then takes a small step forward with his right leg and waves his left arm forward a little. Then he lowers his arm and takes a step back, then turns almost 180 degrees to the right and takes a step forward.",
370
+ "dataset": "snapmogen_original_filtered_test",
371
+ "dataset_index": 1742,
372
+ "generated_matching_distance": 1.342306137084961,
373
+ "generated_reference_distance_gap": 0.1792541742324829,
374
+ "generated_text_rank": 11,
375
+ "reference_matching_distance": 1.163051962852478,
376
+ "reference_text_rank": 1,
377
+ "sample_index": 474
378
+ },
379
+ {
380
+ "caption": "sks person throws an object several times towards a specific direction.",
381
+ "dataset": "permo_original_filtered_test",
382
+ "dataset_index": 256,
383
+ "generated_matching_distance": 1.2616204023361206,
384
+ "generated_reference_distance_gap": 0.09694695472717285,
385
+ "generated_text_rank": 1,
386
+ "reference_matching_distance": 1.1646734476089478,
387
+ "reference_text_rank": 1,
388
+ "sample_index": 436
389
+ },
390
+ {
391
+ "caption": "the man is walking forwards and then backwards and then sideways to complete a circle.",
392
+ "dataset": "humanml_original_filtered_test",
393
+ "dataset_index": 737,
394
+ "generated_matching_distance": 1.384908676147461,
395
+ "generated_reference_distance_gap": 0.21945810317993164,
396
+ "generated_text_rank": 21,
397
+ "reference_matching_distance": 1.1654505729675293,
398
+ "reference_text_rank": 1,
399
+ "sample_index": 174
400
+ },
401
+ {
402
+ "caption": "A person is doing overhead trap raises while standing on the ground.",
403
+ "dataset": "motionllama_original_filtered_test",
404
+ "dataset_index": 85,
405
+ "generated_matching_distance": 1.4425371885299683,
406
+ "generated_reference_distance_gap": 0.2765030860900879,
407
+ "generated_text_rank": 28,
408
+ "reference_matching_distance": 1.1660341024398804,
409
+ "reference_text_rank": 1,
410
+ "sample_index": 345
411
+ },
412
+ {
413
+ "caption": "sks person waves to someone, moving one arm repeatedly.",
414
+ "dataset": "permo_original_filtered_test",
415
+ "dataset_index": 367,
416
+ "generated_matching_distance": 1.3144952058792114,
417
+ "generated_reference_distance_gap": 0.14827477931976318,
418
+ "generated_text_rank": 1,
419
+ "reference_matching_distance": 1.1662204265594482,
420
+ "reference_text_rank": 1,
421
+ "sample_index": 439
422
+ },
423
+ {
424
+ "caption": "A person is lifting a barbell.",
425
+ "dataset": "motionllama_original_filtered_test",
426
+ "dataset_index": 361,
427
+ "generated_matching_distance": 1.4584448337554932,
428
+ "generated_reference_distance_gap": 0.2917494773864746,
429
+ "generated_text_rank": 29,
430
+ "reference_matching_distance": 1.1666953563690186,
431
+ "reference_text_rank": 1,
432
+ "sample_index": 355
433
+ },
434
+ {
435
+ "caption": "The man is doing Doppelpenetration. A person in a purple jacket executes dynamic sword movements with precision.",
436
+ "dataset": "motiongv_original_filtered_test",
437
+ "dataset_index": 95268,
438
+ "generated_matching_distance": 1.3520954847335815,
439
+ "generated_reference_distance_gap": 0.1852174997329712,
440
+ "generated_text_rank": 13,
441
+ "reference_matching_distance": 1.1668779850006104,
442
+ "reference_text_rank": 1,
443
+ "sample_index": 343
444
+ }
445
+ ],
446
+ "largest_generated_reference_distance_gap": [
447
+ {
448
+ "caption": "A person is lifting a barbell.",
449
+ "dataset": "motionllama_original_filtered_test",
450
+ "dataset_index": 361,
451
+ "generated_matching_distance": 1.4584448337554932,
452
+ "generated_reference_distance_gap": 0.2917494773864746,
453
+ "generated_text_rank": 29,
454
+ "reference_matching_distance": 1.1666953563690186,
455
+ "reference_text_rank": 1,
456
+ "sample_index": 355
457
+ },
458
+ {
459
+ "caption": "A person is doing overhead trap raises while standing on the ground.",
460
+ "dataset": "motionllama_original_filtered_test",
461
+ "dataset_index": 85,
462
+ "generated_matching_distance": 1.4425371885299683,
463
+ "generated_reference_distance_gap": 0.2765030860900879,
464
+ "generated_text_rank": 28,
465
+ "reference_matching_distance": 1.1660341024398804,
466
+ "reference_text_rank": 1,
467
+ "sample_index": 345
468
+ },
469
+ {
470
+ "caption": "A martial artist stands firmly with feet wide apart and fists raised, exuding confidence.",
471
+ "dataset": "motiongv_original_filtered_test",
472
+ "dataset_index": 3528,
473
+ "generated_matching_distance": 1.4611055850982666,
474
+ "generated_reference_distance_gap": 0.2699296474456787,
475
+ "generated_text_rank": 31,
476
+ "reference_matching_distance": 1.191175937652588,
477
+ "reference_text_rank": 1,
478
+ "sample_index": 301
479
+ },
480
+ {
481
+ "caption": "The person walks forward in a depressed and sluggish manner, tilting their head to the left and leaning their torso slightly forward, moving their arms gently, and taking short, slow steps.",
482
+ "dataset": "snapmogen_original_filtered_test",
483
+ "dataset_index": 9460,
484
+ "generated_matching_distance": 1.3915009498596191,
485
+ "generated_reference_distance_gap": 0.26295387744903564,
486
+ "generated_text_rank": 18,
487
+ "reference_matching_distance": 1.1285470724105835,
488
+ "reference_text_rank": 1,
489
+ "sample_index": 506
490
+ },
491
+ {
492
+ "caption": "A woman stands still, her head tilted slightly and hand raised in emphasis, conveying engagement and focus.",
493
+ "dataset": "motiongv_original_filtered_test",
494
+ "dataset_index": 55013,
495
+ "generated_matching_distance": 1.465605616569519,
496
+ "generated_reference_distance_gap": 0.2608436346054077,
497
+ "generated_text_rank": 30,
498
+ "reference_matching_distance": 1.2047619819641113,
499
+ "reference_text_rank": 1,
500
+ "sample_index": 325
501
+ },
502
+ {
503
+ "caption": "A person is standing upleft with his legs wide, shifting his body weight to her right foot. She turns to the left and walks forward. Then the person turns to the left, raises her right hand to her face, and lowers his head.",
504
+ "dataset": "snapmogen_original_filtered_test",
505
+ "dataset_index": 6500,
506
+ "generated_matching_distance": 1.4315664768218994,
507
+ "generated_reference_distance_gap": 0.2479323148727417,
508
+ "generated_text_rank": 22,
509
+ "reference_matching_distance": 1.1836341619491577,
510
+ "reference_text_rank": 1,
511
+ "sample_index": 491
512
+ },
513
+ {
514
+ "caption": "The person takes a small step to their left and swings their arms in front of them. Gradually, the person's arms swing less, and after a while, they stop. Then the person slightly raises their arms, bending their elbows.",
515
+ "dataset": "snapmogen_original_filtered_test",
516
+ "dataset_index": 5369,
517
+ "generated_matching_distance": 1.4632925987243652,
518
+ "generated_reference_distance_gap": 0.24439692497253418,
519
+ "generated_text_rank": 30,
520
+ "reference_matching_distance": 1.218895673751831,
521
+ "reference_text_rank": 1,
522
+ "sample_index": 488
523
+ },
524
+ {
525
+ "caption": "The person walks in a circle. He walks in a circle to the right, turning slightly over his right shoulder, while waving his arms back and forth in a relaxed manner.",
526
+ "dataset": "snapmogen_original_filtered_test",
527
+ "dataset_index": 2114,
528
+ "generated_matching_distance": 1.385533094406128,
529
+ "generated_reference_distance_gap": 0.24202275276184082,
530
+ "generated_text_rank": 14,
531
+ "reference_matching_distance": 1.143510341644287,
532
+ "reference_text_rank": 1,
533
+ "sample_index": 476
534
+ },
535
+ {
536
+ "caption": "He places the mouse in his right hand, grasps the keyboard with both, and then sets it down.",
537
+ "dataset": "motionllama_original_filtered_test",
538
+ "dataset_index": 1196,
539
+ "generated_matching_distance": 1.4222255945205688,
540
+ "generated_reference_distance_gap": 0.24091553688049316,
541
+ "generated_text_rank": 18,
542
+ "reference_matching_distance": 1.1813100576400757,
543
+ "reference_text_rank": 1,
544
+ "sample_index": 381
545
+ },
546
+ {
547
+ "caption": "He alternates arms while performing dumbbell curl trifectas.",
548
+ "dataset": "motionllama_original_filtered_test",
549
+ "dataset_index": 213,
550
+ "generated_matching_distance": 1.421977162361145,
551
+ "generated_reference_distance_gap": 0.23986399173736572,
552
+ "generated_text_rank": 21,
553
+ "reference_matching_distance": 1.1821131706237793,
554
+ "reference_text_rank": 1,
555
+ "sample_index": 351
556
+ },
557
+ {
558
+ "caption": "The person extends their arms at chest level. Then, the person slightly shakes their head and rotates their arms in different directions. The person then holds their hands extended at chest level.",
559
+ "dataset": "snapmogen_original_filtered_test",
560
+ "dataset_index": 8764,
561
+ "generated_matching_distance": 1.4769279956817627,
562
+ "generated_reference_distance_gap": 0.23555612564086914,
563
+ "generated_text_rank": 32,
564
+ "reference_matching_distance": 1.2413718700408936,
565
+ "reference_text_rank": 1,
566
+ "sample_index": 502
567
+ },
568
+ {
569
+ "caption": "A person stands with his arms down at his sides and poses. He steps forward with his left foot, turns left, turns right, faces forward, and takes a step backward.",
570
+ "dataset": "snapmogen_original_filtered_test",
571
+ "dataset_index": 11780,
572
+ "generated_matching_distance": 1.4174388647079468,
573
+ "generated_reference_distance_gap": 0.22870850563049316,
574
+ "generated_text_rank": 28,
575
+ "reference_matching_distance": 1.1887303590774536,
576
+ "reference_text_rank": 1,
577
+ "sample_index": 511
578
+ },
579
+ {
580
+ "caption": "A woman in a pink tank top performs a stretching movement, starting with clasped hands behind her back and a slight bend in her arms as she lifts her chest and pulls her shoulders back, exuding a calm and focused demeanor.",
581
+ "dataset": "motiongv_original_filtered_test",
582
+ "dataset_index": 56977,
583
+ "generated_matching_distance": 1.404207468032837,
584
+ "generated_reference_distance_gap": 0.22663724422454834,
585
+ "generated_text_rank": 24,
586
+ "reference_matching_distance": 1.1775702238082886,
587
+ "reference_text_rank": 1,
588
+ "sample_index": 326
589
+ },
590
+ {
591
+ "caption": "The person lunges forward with the right foot, bending the right knee 90 degrees, then stands up and lunges forward with the left foot, bending the left knee 90 degrees, then lunges forward with the right foot and stands standing on the right leg.",
592
+ "dataset": "snapmogen_original_filtered_test",
593
+ "dataset_index": 625,
594
+ "generated_matching_distance": 1.400835394859314,
595
+ "generated_reference_distance_gap": 0.2253406047821045,
596
+ "generated_text_rank": 24,
597
+ "reference_matching_distance": 1.1754947900772095,
598
+ "reference_text_rank": 1,
599
+ "sample_index": 470
600
+ },
601
+ {
602
+ "caption": "A woman performs sign language by raising her right hand and moving it in a circular motion while keeping her left hand still and her body upright.",
603
+ "dataset": "motiongv_original_filtered_test",
604
+ "dataset_index": 51958,
605
+ "generated_matching_distance": 1.472267508506775,
606
+ "generated_reference_distance_gap": 0.22463548183441162,
607
+ "generated_text_rank": 31,
608
+ "reference_matching_distance": 1.2476320266723633,
609
+ "reference_text_rank": 1,
610
+ "sample_index": 322
611
+ },
612
+ {
613
+ "caption": "He holds dumbbells and performs four overhead presses.",
614
+ "dataset": "motionllama_original_filtered_test",
615
+ "dataset_index": 166,
616
+ "generated_matching_distance": 1.4176703691482544,
617
+ "generated_reference_distance_gap": 0.22029900550842285,
618
+ "generated_text_rank": 22,
619
+ "reference_matching_distance": 1.1973713636398315,
620
+ "reference_text_rank": 1,
621
+ "sample_index": 348
622
+ },
623
+ {
624
+ "caption": "Repeated kicks are made by sks person.",
625
+ "dataset": "permo_original_filtered_test",
626
+ "dataset_index": 210,
627
+ "generated_matching_distance": 1.3902548551559448,
628
+ "generated_reference_distance_gap": 0.2194727659225464,
629
+ "generated_text_rank": 12,
630
+ "reference_matching_distance": 1.1707820892333984,
631
+ "reference_text_rank": 1,
632
+ "sample_index": 432
633
+ },
634
+ {
635
+ "caption": "the man is walking forwards and then backwards and then sideways to complete a circle.",
636
+ "dataset": "humanml_original_filtered_test",
637
+ "dataset_index": 737,
638
+ "generated_matching_distance": 1.384908676147461,
639
+ "generated_reference_distance_gap": 0.21945810317993164,
640
+ "generated_text_rank": 21,
641
+ "reference_matching_distance": 1.1654505729675293,
642
+ "reference_text_rank": 1,
643
+ "sample_index": 174
644
+ },
645
+ {
646
+ "caption": "The person stands with their head tilted to the left, then straightens their head and stands while slightly shifting on their feet, tilting their head upward slightly to the left.",
647
+ "dataset": "snapmogen_original_filtered_test",
648
+ "dataset_index": 1383,
649
+ "generated_matching_distance": 1.4192094802856445,
650
+ "generated_reference_distance_gap": 0.2190314531326294,
651
+ "generated_text_rank": 25,
652
+ "reference_matching_distance": 1.2001780271530151,
653
+ "reference_text_rank": 1,
654
+ "sample_index": 473
655
+ },
656
+ {
657
+ "caption": "sks person walks forward, then shifts sideways to alter their path.",
658
+ "dataset": "permo_original_filtered_test",
659
+ "dataset_index": 878,
660
+ "generated_matching_distance": 1.3643834590911865,
661
+ "generated_reference_distance_gap": 0.21602392196655273,
662
+ "generated_text_rank": 15,
663
+ "reference_matching_distance": 1.1483595371246338,
664
+ "reference_text_rank": 1,
665
+ "sample_index": 458
666
+ }
667
+ ],
668
+ "num_samples": 512,
669
+ "worst_generated": [
670
+ {
671
+ "caption": "The person extends their arms at chest level. Then, the person slightly shakes their head and rotates their arms in different directions. The person then holds their hands extended at chest level.",
672
+ "dataset": "snapmogen_original_filtered_test",
673
+ "dataset_index": 8764,
674
+ "generated_matching_distance": 1.4769279956817627,
675
+ "generated_reference_distance_gap": 0.23555612564086914,
676
+ "generated_text_rank": 32,
677
+ "reference_matching_distance": 1.2413718700408936,
678
+ "reference_text_rank": 1,
679
+ "sample_index": 502
680
+ },
681
+ {
682
+ "caption": "A woman performs sign language by raising her right hand and moving it in a circular motion while keeping her left hand still and her body upright.",
683
+ "dataset": "motiongv_original_filtered_test",
684
+ "dataset_index": 51958,
685
+ "generated_matching_distance": 1.472267508506775,
686
+ "generated_reference_distance_gap": 0.22463548183441162,
687
+ "generated_text_rank": 31,
688
+ "reference_matching_distance": 1.2476320266723633,
689
+ "reference_text_rank": 1,
690
+ "sample_index": 322
691
+ },
692
+ {
693
+ "caption": "A martial artist stands firmly with feet wide apart and fists raised, exuding confidence.",
694
+ "dataset": "motiongv_original_filtered_test",
695
+ "dataset_index": 3528,
696
+ "generated_matching_distance": 1.4611055850982666,
697
+ "generated_reference_distance_gap": 0.2699296474456787,
698
+ "generated_text_rank": 31,
699
+ "reference_matching_distance": 1.191175937652588,
700
+ "reference_text_rank": 1,
701
+ "sample_index": 301
702
+ },
703
+ {
704
+ "caption": "A woman stands still, her head tilted slightly and hand raised in emphasis, conveying engagement and focus.",
705
+ "dataset": "motiongv_original_filtered_test",
706
+ "dataset_index": 55013,
707
+ "generated_matching_distance": 1.465605616569519,
708
+ "generated_reference_distance_gap": 0.2608436346054077,
709
+ "generated_text_rank": 30,
710
+ "reference_matching_distance": 1.2047619819641113,
711
+ "reference_text_rank": 1,
712
+ "sample_index": 325
713
+ },
714
+ {
715
+ "caption": "The person takes a small step to their left and swings their arms in front of them. Gradually, the person's arms swing less, and after a while, they stop. Then the person slightly raises their arms, bending their elbows.",
716
+ "dataset": "snapmogen_original_filtered_test",
717
+ "dataset_index": 5369,
718
+ "generated_matching_distance": 1.4632925987243652,
719
+ "generated_reference_distance_gap": 0.24439692497253418,
720
+ "generated_text_rank": 30,
721
+ "reference_matching_distance": 1.218895673751831,
722
+ "reference_text_rank": 1,
723
+ "sample_index": 488
724
+ },
725
+ {
726
+ "caption": "She performs ballet with a strong sense of revulsion.",
727
+ "dataset": "fitness_original_filtered_test",
728
+ "dataset_index": 2,
729
+ "generated_matching_distance": 1.460925579071045,
730
+ "generated_reference_distance_gap": 0.16499972343444824,
731
+ "generated_text_rank": 30,
732
+ "reference_matching_distance": 1.2959258556365967,
733
+ "reference_text_rank": 1,
734
+ "sample_index": 129
735
+ },
736
+ {
737
+ "caption": "the person is standing with his hands on his side.",
738
+ "dataset": "humanml_original_filtered_test",
739
+ "dataset_index": 1365,
740
+ "generated_matching_distance": 1.430628776550293,
741
+ "generated_reference_distance_gap": 0.15506017208099365,
742
+ "generated_text_rank": 30,
743
+ "reference_matching_distance": 1.2755686044692993,
744
+ "reference_text_rank": 1,
745
+ "sample_index": 178
746
+ },
747
+ {
748
+ "caption": "A person sits down, then uses his right hand to pick up the pen.",
749
+ "dataset": "motionllama_original_filtered_test",
750
+ "dataset_index": 786,
751
+ "generated_matching_distance": 1.411001443862915,
752
+ "generated_reference_distance_gap": 0.17560267448425293,
753
+ "generated_text_rank": 30,
754
+ "reference_matching_distance": 1.235398769378662,
755
+ "reference_text_rank": 1,
756
+ "sample_index": 369
757
+ },
758
+ {
759
+ "caption": "The person stands calmly upleft, leans slightly to the right and left, and lifts their head up.",
760
+ "dataset": "snapmogen_original_filtered_test",
761
+ "dataset_index": 8732,
762
+ "generated_matching_distance": 1.4056349992752075,
763
+ "generated_reference_distance_gap": 0.19578826427459717,
764
+ "generated_text_rank": 30,
765
+ "reference_matching_distance": 1.2098467350006104,
766
+ "reference_text_rank": 1,
767
+ "sample_index": 501
768
+ },
769
+ {
770
+ "caption": "A person is lifting a barbell.",
771
+ "dataset": "motionllama_original_filtered_test",
772
+ "dataset_index": 361,
773
+ "generated_matching_distance": 1.4584448337554932,
774
+ "generated_reference_distance_gap": 0.2917494773864746,
775
+ "generated_text_rank": 29,
776
+ "reference_matching_distance": 1.1666953563690186,
777
+ "reference_text_rank": 1,
778
+ "sample_index": 355
779
+ },
780
+ {
781
+ "caption": "She performs ballet with a strong sense of revulsion.",
782
+ "dataset": "fitness_original_filtered_test",
783
+ "dataset_index": 3,
784
+ "generated_matching_distance": 1.4565238952636719,
785
+ "generated_reference_distance_gap": 0.1704561710357666,
786
+ "generated_text_rank": 29,
787
+ "reference_matching_distance": 1.2860677242279053,
788
+ "reference_text_rank": 1,
789
+ "sample_index": 130
790
+ },
791
+ {
792
+ "caption": "A woman exercises with high energy. Initially, she extends her arms while standing still. Next, she swings her arms in and out while lightly bouncing on her feet, exuding a lively and cheerful demeanor throughout, making her movements seem smooth and rhythmic.",
793
+ "dataset": "motiongv_original_filtered_test",
794
+ "dataset_index": 30180,
795
+ "generated_matching_distance": 1.4433854818344116,
796
+ "generated_reference_distance_gap": 0.1368476152420044,
797
+ "generated_text_rank": 29,
798
+ "reference_matching_distance": 1.3065378665924072,
799
+ "reference_text_rank": 2,
800
+ "sample_index": 314
801
+ },
802
+ {
803
+ "caption": "The person calmly warms up the left hand. The person turns to the right while standing upleft and makes two full turns and in the process raises and lowers the left arm four times while the right arm is lowered.",
804
+ "dataset": "snapmogen_original_filtered_test",
805
+ "dataset_index": 7158,
806
+ "generated_matching_distance": 1.4316307306289673,
807
+ "generated_reference_distance_gap": 0.17985808849334717,
808
+ "generated_text_rank": 29,
809
+ "reference_matching_distance": 1.2517726421356201,
810
+ "reference_text_rank": 2,
811
+ "sample_index": 494
812
+ },
813
+ {
814
+ "caption": "Sport fitness enthusiasts engage in left and right arm lifts while standing.",
815
+ "dataset": "fitness_original_filtered_test",
816
+ "dataset_index": 46,
817
+ "generated_matching_distance": 1.4308847188949585,
818
+ "generated_reference_distance_gap": 0.1247556209564209,
819
+ "generated_text_rank": 29,
820
+ "reference_matching_distance": 1.3061290979385376,
821
+ "reference_text_rank": 2,
822
+ "sample_index": 145
823
+ },
824
+ {
825
+ "caption": "Sport fitness involves lifting arms left and right while standing.",
826
+ "dataset": "fitness_original_filtered_test",
827
+ "dataset_index": 53,
828
+ "generated_matching_distance": 1.4219019412994385,
829
+ "generated_reference_distance_gap": 0.16597938537597656,
830
+ "generated_text_rank": 29,
831
+ "reference_matching_distance": 1.255922555923462,
832
+ "reference_text_rank": 1,
833
+ "sample_index": 150
834
+ },
835
+ {
836
+ "caption": "The person stands with their feet wide apart and torso bent backwards. The person then turns to the left, walks forward, bends down and lowers and raises their arms alternately. The person then stops, turns slightly to the right, and straightens up.",
837
+ "dataset": "snapmogen_original_filtered_test",
838
+ "dataset_index": 9620,
839
+ "generated_matching_distance": 1.4140336513519287,
840
+ "generated_reference_distance_gap": 0.16541707515716553,
841
+ "generated_text_rank": 29,
842
+ "reference_matching_distance": 1.2486165761947632,
843
+ "reference_text_rank": 1,
844
+ "sample_index": 507
845
+ },
846
+ {
847
+ "caption": "Sport fitness involves lifting arms left and right while standing.",
848
+ "dataset": "fitness_original_filtered_test",
849
+ "dataset_index": 58,
850
+ "generated_matching_distance": 1.4433379173278809,
851
+ "generated_reference_distance_gap": 0.1963796615600586,
852
+ "generated_text_rank": 28,
853
+ "reference_matching_distance": 1.2469582557678223,
854
+ "reference_text_rank": 1,
855
+ "sample_index": 153
856
+ },
857
+ {
858
+ "caption": "A person is doing overhead trap raises while standing on the ground.",
859
+ "dataset": "motionllama_original_filtered_test",
860
+ "dataset_index": 85,
861
+ "generated_matching_distance": 1.4425371885299683,
862
+ "generated_reference_distance_gap": 0.2765030860900879,
863
+ "generated_text_rank": 28,
864
+ "reference_matching_distance": 1.1660341024398804,
865
+ "reference_text_rank": 1,
866
+ "sample_index": 345
867
+ },
868
+ {
869
+ "caption": "Sport fitness involves lifting arms left and right while standing.",
870
+ "dataset": "fitness_original_filtered_test",
871
+ "dataset_index": 57,
872
+ "generated_matching_distance": 1.4382672309875488,
873
+ "generated_reference_distance_gap": 0.1747530698776245,
874
+ "generated_text_rank": 28,
875
+ "reference_matching_distance": 1.2635141611099243,
876
+ "reference_text_rank": 1,
877
+ "sample_index": 152
878
+ },
879
+ {
880
+ "caption": "She performs ballet with a strong sense of revulsion.",
881
+ "dataset": "fitness_original_filtered_test",
882
+ "dataset_index": 6,
883
+ "generated_matching_distance": 1.4360405206680298,
884
+ "generated_reference_distance_gap": 0.16289913654327393,
885
+ "generated_text_rank": 28,
886
+ "reference_matching_distance": 1.2731413841247559,
887
+ "reference_text_rank": 1,
888
+ "sample_index": 131
889
+ }
890
+ ],
891
+ "worst_reference": [
892
+ {
893
+ "caption": "lift legs alternately, bend knees, extend arms forward to maintain balance; style: balanced",
894
+ "dataset": "amass_original_filtered_test",
895
+ "dataset_index": 2739,
896
+ "generated_matching_distance": 1.3685790300369263,
897
+ "generated_reference_distance_gap": -0.04497110843658447,
898
+ "generated_text_rank": 11,
899
+ "reference_matching_distance": 1.4135501384735107,
900
+ "reference_text_rank": 28,
901
+ "sample_index": 61
902
+ },
903
+ {
904
+ "caption": "step in place with bent knees, move arms in sync; style: rhythmic",
905
+ "dataset": "omomo_original_filtered_test",
906
+ "dataset_index": 934,
907
+ "generated_matching_distance": 1.2867180109024048,
908
+ "generated_reference_distance_gap": -0.08043253421783447,
909
+ "generated_text_rank": 2,
910
+ "reference_matching_distance": 1.3671505451202393,
911
+ "reference_text_rank": 24,
912
+ "sample_index": 394
913
+ },
914
+ {
915
+ "caption": "The right leg stretch is a part of sport fitness.",
916
+ "dataset": "fitness_original_filtered_test",
917
+ "dataset_index": 95,
918
+ "generated_matching_distance": 1.4103211164474487,
919
+ "generated_reference_distance_gap": 0.028392672538757324,
920
+ "generated_text_rank": 21,
921
+ "reference_matching_distance": 1.3819284439086914,
922
+ "reference_text_rank": 20,
923
+ "sample_index": 171
924
+ },
925
+ {
926
+ "caption": "Someone sits at a table, holds a pen, and writes on a document.; style: controlled",
927
+ "dataset": "idea400_original_filtered_test",
928
+ "dataset_index": 840,
929
+ "generated_matching_distance": 1.3414204120635986,
930
+ "generated_reference_distance_gap": -0.051953792572021484,
931
+ "generated_text_rank": 4,
932
+ "reference_matching_distance": 1.3933742046356201,
933
+ "reference_text_rank": 16,
934
+ "sample_index": 230
935
+ },
936
+ {
937
+ "caption": "stand motionless in upright position; style: static",
938
+ "dataset": "100style_original_filtered_test",
939
+ "dataset_index": 2007,
940
+ "generated_matching_distance": 1.3318371772766113,
941
+ "generated_reference_distance_gap": -0.02785325050354004,
942
+ "generated_text_rank": 8,
943
+ "reference_matching_distance": 1.3596904277801514,
944
+ "reference_text_rank": 12,
945
+ "sample_index": 28
946
+ },
947
+ {
948
+ "caption": "He begins by standing upright with relaxed arms.",
949
+ "dataset": "motiongv_original_filtered_test",
950
+ "dataset_index": 10692,
951
+ "generated_matching_distance": 1.4310532808303833,
952
+ "generated_reference_distance_gap": 0.07549643516540527,
953
+ "generated_text_rank": 27,
954
+ "reference_matching_distance": 1.355556845664978,
955
+ "reference_text_rank": 12,
956
+ "sample_index": 303
957
+ },
958
+ {
959
+ "caption": "The right leg stretch is a part of sport fitness.",
960
+ "dataset": "fitness_original_filtered_test",
961
+ "dataset_index": 94,
962
+ "generated_matching_distance": 1.2829440832138062,
963
+ "generated_reference_distance_gap": -0.06328606605529785,
964
+ "generated_text_rank": 1,
965
+ "reference_matching_distance": 1.346230149269104,
966
+ "reference_text_rank": 11,
967
+ "sample_index": 170
968
+ },
969
+ {
970
+ "caption": "perform small jumps and cross steps with extended arms; style: dynamic",
971
+ "dataset": "amass_original_filtered_test",
972
+ "dataset_index": 6011,
973
+ "generated_matching_distance": 1.3225088119506836,
974
+ "generated_reference_distance_gap": 0.024164915084838867,
975
+ "generated_text_rank": 1,
976
+ "reference_matching_distance": 1.2983438968658447,
977
+ "reference_text_rank": 10,
978
+ "sample_index": 84
979
+ },
980
+ {
981
+ "caption": "turn right, extend arms wide, step backward smoothly; style: smooth",
982
+ "dataset": "100style_original_filtered_test",
983
+ "dataset_index": 973,
984
+ "generated_matching_distance": 1.3289071321487427,
985
+ "generated_reference_distance_gap": 0.0061490535736083984,
986
+ "generated_text_rank": 6,
987
+ "reference_matching_distance": 1.3227580785751343,
988
+ "reference_text_rank": 9,
989
+ "sample_index": 15
990
+ },
991
+ {
992
+ "caption": "walk to the right, swing arms in sync with stepping motion; style: steady, natural",
993
+ "dataset": "omomo_original_filtered_test",
994
+ "dataset_index": 1184,
995
+ "generated_matching_distance": 1.2476807832717896,
996
+ "generated_reference_distance_gap": -0.06195378303527832,
997
+ "generated_text_rank": 1,
998
+ "reference_matching_distance": 1.3096345663070679,
999
+ "reference_text_rank": 8,
1000
+ "sample_index": 409
1001
+ },
1002
+ {
1003
+ "caption": "perform continuous balanced dance with stepping and arm motions; style: fluid, balanced",
1004
+ "dataset": "100style_original_filtered_test",
1005
+ "dataset_index": 277,
1006
+ "generated_matching_distance": 1.3593848943710327,
1007
+ "generated_reference_distance_gap": 0.05692768096923828,
1008
+ "generated_text_rank": 7,
1009
+ "reference_matching_distance": 1.3024572134017944,
1010
+ "reference_text_rank": 8,
1011
+ "sample_index": 9
1012
+ },
1013
+ {
1014
+ "caption": "A person stands on the ground, practicing overhead trap raises.",
1015
+ "dataset": "motionllama_original_filtered_test",
1016
+ "dataset_index": 173,
1017
+ "generated_matching_distance": 1.4499162435531616,
1018
+ "generated_reference_distance_gap": 0.10569095611572266,
1019
+ "generated_text_rank": 27,
1020
+ "reference_matching_distance": 1.344225287437439,
1021
+ "reference_text_rank": 7,
1022
+ "sample_index": 349
1023
+ },
1024
+ {
1025
+ "caption": "bend knees, lower body, and sit down on the ground; style: controlled",
1026
+ "dataset": "amass_original_filtered_test",
1027
+ "dataset_index": 3877,
1028
+ "generated_matching_distance": 1.2690072059631348,
1029
+ "generated_reference_distance_gap": -0.07488059997558594,
1030
+ "generated_text_rank": 1,
1031
+ "reference_matching_distance": 1.3438878059387207,
1032
+ "reference_text_rank": 7,
1033
+ "sample_index": 66
1034
+ },
1035
+ {
1036
+ "caption": "perform repetitive side steps and arm movements; style: rhythmic",
1037
+ "dataset": "amass_original_filtered_test",
1038
+ "dataset_index": 5753,
1039
+ "generated_matching_distance": 1.299026608467102,
1040
+ "generated_reference_distance_gap": -0.03400909900665283,
1041
+ "generated_text_rank": 3,
1042
+ "reference_matching_distance": 1.3330357074737549,
1043
+ "reference_text_rank": 7,
1044
+ "sample_index": 83
1045
+ },
1046
+ {
1047
+ "caption": "turn torso, bend knees, lower arms to sides of body; style: controlled",
1048
+ "dataset": "amass_original_filtered_test",
1049
+ "dataset_index": 2767,
1050
+ "generated_matching_distance": 1.3642807006835938,
1051
+ "generated_reference_distance_gap": 0.04586946964263916,
1052
+ "generated_text_rank": 8,
1053
+ "reference_matching_distance": 1.3184112310409546,
1054
+ "reference_text_rank": 7,
1055
+ "sample_index": 62
1056
+ },
1057
+ {
1058
+ "caption": "A man walks across a space, alternating left and right foot steps in a bipedal pattern, while swinging its arms gently in a rhythm opposite to its leg movements to maintain balance.; style: rhythmic",
1059
+ "dataset": "idea400_original_filtered_test",
1060
+ "dataset_index": 1817,
1061
+ "generated_matching_distance": 1.4023194313049316,
1062
+ "generated_reference_distance_gap": 0.09062647819519043,
1063
+ "generated_text_rank": 18,
1064
+ "reference_matching_distance": 1.3116929531097412,
1065
+ "reference_text_rank": 7,
1066
+ "sample_index": 240
1067
+ },
1068
+ {
1069
+ "caption": "swing shoulders, rotate torso, and step in circular patterns; style: energetic rhythmic",
1070
+ "dataset": "100style_original_filtered_test",
1071
+ "dataset_index": 2683,
1072
+ "generated_matching_distance": 1.2990708351135254,
1073
+ "generated_reference_distance_gap": -0.002228975296020508,
1074
+ "generated_text_rank": 3,
1075
+ "reference_matching_distance": 1.301299810409546,
1076
+ "reference_text_rank": 7,
1077
+ "sample_index": 36
1078
+ },
1079
+ {
1080
+ "caption": "sprint forward, execute 180 turn, stop; style: agile",
1081
+ "dataset": "lafan1_original_filtered_test",
1082
+ "dataset_index": 157,
1083
+ "generated_matching_distance": 1.2624266147613525,
1084
+ "generated_reference_distance_gap": 0.01119387149810791,
1085
+ "generated_text_rank": 1,
1086
+ "reference_matching_distance": 1.2512327432632446,
1087
+ "reference_text_rank": 7,
1088
+ "sample_index": 276
1089
+ },
1090
+ {
1091
+ "caption": "A young woman exercises with great energy.",
1092
+ "dataset": "motiongv_original_filtered_test",
1093
+ "dataset_index": 70451,
1094
+ "generated_matching_distance": 1.3851017951965332,
1095
+ "generated_reference_distance_gap": 0.04356992244720459,
1096
+ "generated_text_rank": 15,
1097
+ "reference_matching_distance": 1.3415318727493286,
1098
+ "reference_text_rank": 6,
1099
+ "sample_index": 334
1100
+ },
1101
+ {
1102
+ "caption": "Shift body sideways and take small forward steps; style: mechanical, deliberate",
1103
+ "dataset": "omomo_original_filtered_test",
1104
+ "dataset_index": 1189,
1105
+ "generated_matching_distance": 1.3251229524612427,
1106
+ "generated_reference_distance_gap": 0.00553286075592041,
1107
+ "generated_text_rank": 3,
1108
+ "reference_matching_distance": 1.3195900917053223,
1109
+ "reference_text_rank": 6,
1110
+ "sample_index": 410
1111
+ }
1112
+ ]
1113
+ }
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/summary.md ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ # Generation Benchmark Summary
2
+
3
+ - benchmark: artifact_text
4
+ - label: Kimodo-G1-SEED-v1
5
+ - exp: g1_transformer_guided_300m
6
+ - split: test
7
+
8
+ | ckpt | FID(seq) | KID(seq) | R@1 | R@2 | R@3 | Diversity | Matching | MultiModal Dist | BeatAlign | BeatAlign Ref | BeatAlign Gap | FIDk | FIDg | Divk | Divg | PFC | PFC Ref | contact_sliding | foot_ground_error | body_jerk | PJ | AUJ | Tracker g-MPJPE | Tracker MPJPE | Tracker E_vel | Tracker E_acc |
9
+ |---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
10
+ | Kimodo-G1-SEED-v1 | **0.5887** | **0.0009 +/- 0.0000** | **0.2402** | **0.3398** | **0.4121** | **1.1859** | **1.3312** | **1.3312 +/- 0.0605** | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | **0.3516 +/- 0.4755** | **0.0099 +/- 0.0653** | **72.1549 +/- 79.0538** | **0.0578 +/- 0.1226** | **1.8406 +/- 0.3404** | N/A | N/A | N/A | N/A |
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/tracker_executed.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "enabled": false
3
+ }
benchmark/results/mixed_modalities_all_v1/text/kimodo_g1_seed_tracker_120f_transition/transition_metrics.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auj": {
3
+ "max": 6.514222457101631,
4
+ "mean": 1.8406120086684412,
5
+ "min": 1.4860971376024386,
6
+ "std": 0.3403637125100069
7
+ },
8
+ "body_auj": {
9
+ "max": 6.514222457101631,
10
+ "mean": 1.8406120086684412,
11
+ "min": 1.4860971376024386,
12
+ "std": 0.3403637125100069
13
+ },
14
+ "body_pj": {
15
+ "max": 1.9185047969222069,
16
+ "mean": 0.0578470742943864,
17
+ "min": 0.0004047006368637085,
18
+ "std": 0.12257751816949211
19
+ },
20
+ "boundary_frames": [
21
+ 60
22
+ ],
23
+ "chunk_length": 60,
24
+ "definition": "PJ=max L1 jerk over links/time; AUJ=sum_t max_link |jerk(t)-reference_jerk_level| over transition windows.",
25
+ "enabled": true,
26
+ "num_frames": 120,
27
+ "num_samples": 512,
28
+ "pj": {
29
+ "max": 1.9185047969222069,
30
+ "mean": 0.0578470742943864,
31
+ "min": 0.0004047006368637085,
32
+ "std": 0.12257751816949211
33
+ },
34
+ "qpos_auj": {
35
+ "max": 144.49321365365512,
36
+ "mean": 6.459746310243598,
37
+ "min": 4.2940439462623115,
38
+ "std": 8.748215194373627
39
+ },
40
+ "qpos_pj": {
41
+ "max": 12.888956546783447,
42
+ "mean": 0.706420054541411,
43
+ "min": 0.0026228129863739014,
44
+ "std": 1.8797036959973208
45
+ },
46
+ "reference_body_jerk_level": 0.032168975994990016,
47
+ "reference_qpos_jerk_level": 0.07551584170589375,
48
+ "transition_window_length": 60,
49
+ "transition_windows": [
50
+ [
51
+ 30,
52
+ 90
53
+ ]
54
+ ]
55
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/artifact_samples.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/benchmark.json ADDED
@@ -0,0 +1,265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark": "artifact_text",
3
+ "dataset_metrics": {
4
+ "filtered_original_mixed_modalities_all_materialized_test": {
5
+ "embedding": {
6
+ "diversity_generated": 1.3626155066538168,
7
+ "motion_fid": 0.10886979363854854,
8
+ "motion_kid": {
9
+ "mean": 6.802706844455386e-05,
10
+ "std": 0.0
11
+ },
12
+ "num_samples": 1024
13
+ },
14
+ "num_samples": 1024,
15
+ "physical": {
16
+ "body_jerk_mean": {
17
+ "max": 360.24395751953125,
18
+ "mean": 53.740967999212444,
19
+ "min": 1.9295663833618164,
20
+ "std": 55.003332467755946
21
+ },
22
+ "contact_sliding_speed": {
23
+ "max": 3.483940601348877,
24
+ "mean": 0.4634235018760364,
25
+ "min": 0.0,
26
+ "std": 0.573620541786601
27
+ },
28
+ "foot_ground_error": {
29
+ "max": 0.2830135226249695,
30
+ "mean": 0.04558111644996643,
31
+ "min": 0.001008824910968542,
32
+ "std": 0.029560935890020977
33
+ },
34
+ "num_samples": 1024
35
+ },
36
+ "text": {
37
+ "generated": {
38
+ "matching_distance": {
39
+ "max": 1.5298819541931152,
40
+ "mean": 1.158189934794791,
41
+ "min": 0.8833644986152649,
42
+ "std": 0.1090334975205812
43
+ },
44
+ "matching_score": 1.1581900119781494,
45
+ "mean_text_rank": 6.603515625,
46
+ "median_text_rank": 4.0,
47
+ "r_precision": [
48
+ 0.30078125,
49
+ 0.4052734375,
50
+ 0.4970703125
51
+ ],
52
+ "retrieval_batch_size": 32,
53
+ "retrieval_num_batches": 32,
54
+ "retrieval_num_samples": 1024,
55
+ "retrieval_order_policy": "manifest_order",
56
+ "retrieval_unique_datasets_per_batch": {
57
+ "max": 1.0,
58
+ "mean": 1.0,
59
+ "min": 1.0,
60
+ "std": 0.0
61
+ }
62
+ },
63
+ "reference": {
64
+ "matching_distance": {
65
+ "max": 1.524587631225586,
66
+ "mean": 1.1037589554325677,
67
+ "min": 0.8309325575828552,
68
+ "std": 0.10401542255778194
69
+ },
70
+ "matching_score": 1.1037590503692627,
71
+ "mean_text_rank": 5.2646484375,
72
+ "median_text_rank": 3.0,
73
+ "r_precision": [
74
+ 0.361328125,
75
+ 0.4873046875,
76
+ 0.576171875
77
+ ],
78
+ "retrieval_batch_size": 32,
79
+ "retrieval_num_batches": 32,
80
+ "retrieval_num_samples": 1024,
81
+ "retrieval_order_policy": "manifest_order",
82
+ "retrieval_unique_datasets_per_batch": {
83
+ "max": 1.0,
84
+ "mean": 1.0,
85
+ "min": 1.0,
86
+ "std": 0.0
87
+ }
88
+ }
89
+ }
90
+ }
91
+ },
92
+ "embedding_frames": 60,
93
+ "embedding_metrics": {
94
+ "diversity_generated": 1.3626155066538168,
95
+ "embedding_frames": 60,
96
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
97
+ "matching_score_generated": 1.1581900119781494,
98
+ "matching_score_reference": 1.1037590503692627,
99
+ "motion_fid": 0.10886979363854854,
100
+ "motion_key": "body_pos_local",
101
+ "motion_kid": {
102
+ "mean": 6.802706844455386e-05,
103
+ "std": 0.0
104
+ },
105
+ "multimodality": null,
106
+ "num_samples": 1024,
107
+ "r_precision_generated": [
108
+ 0.5107421875,
109
+ 0.65625,
110
+ 0.7451171875
111
+ ],
112
+ "r_precision_reference": [
113
+ 0.6396484375,
114
+ 0.8076171875,
115
+ 0.8818359375
116
+ ],
117
+ "text_cache_dir": "${OMG_REPO_ROOT}/outputs/evaluator_text_cache/filtered_original_g1_tmr_bodypos_60f",
118
+ "text_encoder_model": "${OMG_REPO_ROOT}/models/t5-3b-local",
119
+ "text_generated": {
120
+ "matching_distance": {
121
+ "max": 1.5298819541931152,
122
+ "mean": 1.158189934794791,
123
+ "min": 0.8833644986152649,
124
+ "std": 0.1090334975205812
125
+ },
126
+ "matching_score": 1.1581900119781494,
127
+ "mean_text_rank": 3.1650390625,
128
+ "median_text_rank": 1.0,
129
+ "r_precision": [
130
+ 0.5107421875,
131
+ 0.65625,
132
+ 0.7451171875
133
+ ],
134
+ "retrieval_batch_size": 32,
135
+ "retrieval_num_batches": 32,
136
+ "retrieval_num_samples": 1024,
137
+ "retrieval_order_policy": "stratified_by_dataset",
138
+ "retrieval_unique_datasets_per_batch": {
139
+ "max": 1.0,
140
+ "mean": 1.0,
141
+ "min": 1.0,
142
+ "std": 0.0
143
+ }
144
+ },
145
+ "text_reference": {
146
+ "matching_distance": {
147
+ "max": 1.524587631225586,
148
+ "mean": 1.1037589554325677,
149
+ "min": 0.8309325575828552,
150
+ "std": 0.10401542255778194
151
+ },
152
+ "matching_score": 1.1037590503692627,
153
+ "mean_text_rank": 2.029296875,
154
+ "median_text_rank": 1.0,
155
+ "r_precision": [
156
+ 0.6396484375,
157
+ 0.8076171875,
158
+ 0.8818359375
159
+ ],
160
+ "retrieval_batch_size": 32,
161
+ "retrieval_num_batches": 32,
162
+ "retrieval_num_samples": 1024,
163
+ "retrieval_order_policy": "stratified_by_dataset",
164
+ "retrieval_unique_datasets_per_batch": {
165
+ "max": 1.0,
166
+ "mean": 1.0,
167
+ "min": 1.0,
168
+ "std": 0.0
169
+ }
170
+ },
171
+ "text_retrieval_batch_size": 32
172
+ },
173
+ "exp": "g1_transformer_guided_300m",
174
+ "generated_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000/generated_qpos.npz",
175
+ "label": "OmniMotion 300M mixed-modal s045000",
176
+ "num_frames": 60,
177
+ "num_samples": 1024,
178
+ "physical_metrics": {
179
+ "body_jerk_mean": {
180
+ "max": 360.24395751953125,
181
+ "mean": 53.740967999212444,
182
+ "min": 1.9295663833618164,
183
+ "std": 55.003332467755946
184
+ },
185
+ "contact_height_threshold": 0.12,
186
+ "contact_penetration_tolerance": 0.02,
187
+ "contact_sliding_speed": {
188
+ "max": 3.483940601348877,
189
+ "mean": 0.4634235018760364,
190
+ "min": 0.0,
191
+ "std": 0.573620541786601
192
+ },
193
+ "foot_ground_error": {
194
+ "max": 0.2830135226249695,
195
+ "mean": 0.04558111644996643,
196
+ "min": 0.001008824910968542,
197
+ "std": 0.029560935890020977
198
+ },
199
+ "frames": 60,
200
+ "num_samples": 1024
201
+ },
202
+ "real_motion_metrics": {
203
+ "diversity_real": 1.3814387730607542,
204
+ "matching_distance_real": {
205
+ "max": 1.524587631225586,
206
+ "mean": 1.1037589554325677,
207
+ "min": 0.8309325575828552,
208
+ "std": 0.10401542255778194
209
+ },
210
+ "matching_score_real": 1.1037590503692627,
211
+ "motion_fid_reference_eval": 0.0,
212
+ "motion_fid_split_half": 0.0,
213
+ "motion_kid_reference_eval": {
214
+ "mean": -1.0798661567168466e-05,
215
+ "std": 0.0
216
+ },
217
+ "motion_kid_split_half": {
218
+ "mean": -1.0798661567168466e-05,
219
+ "std": 0.0
220
+ },
221
+ "multimodality": "-",
222
+ "num_reference_samples": 1024,
223
+ "num_samples": 1024,
224
+ "r_precision_real": [
225
+ 0.361328125,
226
+ 0.4873046875,
227
+ 0.576171875
228
+ ],
229
+ "text_real": {
230
+ "matching_distance": {
231
+ "max": 1.524587631225586,
232
+ "mean": 1.1037589554325677,
233
+ "min": 0.8309325575828552,
234
+ "std": 0.10401542255778194
235
+ },
236
+ "matching_score": 1.1037590503692627,
237
+ "mean_text_rank": 5.2646484375,
238
+ "median_text_rank": 3.0,
239
+ "r_precision": [
240
+ 0.361328125,
241
+ 0.4873046875,
242
+ 0.576171875
243
+ ],
244
+ "retrieval_batch_size": 32,
245
+ "retrieval_num_batches": 32,
246
+ "retrieval_num_samples": 1024,
247
+ "retrieval_order_policy": "manifest_order",
248
+ "retrieval_unique_datasets_per_batch": null
249
+ }
250
+ },
251
+ "reference_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/reference_qpos.npz",
252
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/sample_metrics.jsonl",
253
+ "sample_rankings_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/sample_rankings.json",
254
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000/samples.jsonl",
255
+ "split": "test",
256
+ "tracker_executed": {
257
+ "enabled": false
258
+ },
259
+ "transition_metrics": {
260
+ "chunk_length": 60,
261
+ "enabled": false,
262
+ "num_frames": 60,
263
+ "reason": "num_frames must be greater than chunk_length"
264
+ }
265
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/dataset_metrics.json ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "filtered_original_mixed_modalities_all_materialized_test": {
3
+ "embedding": {
4
+ "diversity_generated": 1.3626155066538168,
5
+ "motion_fid": 0.10886979363854854,
6
+ "motion_kid": {
7
+ "mean": 6.802706844455386e-05,
8
+ "std": 0.0
9
+ },
10
+ "num_samples": 1024
11
+ },
12
+ "num_samples": 1024,
13
+ "physical": {
14
+ "body_jerk_mean": {
15
+ "max": 360.24395751953125,
16
+ "mean": 53.740967999212444,
17
+ "min": 1.9295663833618164,
18
+ "std": 55.003332467755946
19
+ },
20
+ "contact_sliding_speed": {
21
+ "max": 3.483940601348877,
22
+ "mean": 0.4634235018760364,
23
+ "min": 0.0,
24
+ "std": 0.573620541786601
25
+ },
26
+ "foot_ground_error": {
27
+ "max": 0.2830135226249695,
28
+ "mean": 0.04558111644996643,
29
+ "min": 0.001008824910968542,
30
+ "std": 0.029560935890020977
31
+ },
32
+ "num_samples": 1024
33
+ },
34
+ "text": {
35
+ "generated": {
36
+ "matching_distance": {
37
+ "max": 1.5298819541931152,
38
+ "mean": 1.158189934794791,
39
+ "min": 0.8833644986152649,
40
+ "std": 0.1090334975205812
41
+ },
42
+ "matching_score": 1.1581900119781494,
43
+ "mean_text_rank": 6.603515625,
44
+ "median_text_rank": 4.0,
45
+ "r_precision": [
46
+ 0.30078125,
47
+ 0.4052734375,
48
+ 0.4970703125
49
+ ],
50
+ "retrieval_batch_size": 32,
51
+ "retrieval_num_batches": 32,
52
+ "retrieval_num_samples": 1024,
53
+ "retrieval_order_policy": "manifest_order",
54
+ "retrieval_unique_datasets_per_batch": {
55
+ "max": 1.0,
56
+ "mean": 1.0,
57
+ "min": 1.0,
58
+ "std": 0.0
59
+ }
60
+ },
61
+ "reference": {
62
+ "matching_distance": {
63
+ "max": 1.524587631225586,
64
+ "mean": 1.1037589554325677,
65
+ "min": 0.8309325575828552,
66
+ "std": 0.10401542255778194
67
+ },
68
+ "matching_score": 1.1037590503692627,
69
+ "mean_text_rank": 5.2646484375,
70
+ "median_text_rank": 3.0,
71
+ "r_precision": [
72
+ 0.361328125,
73
+ 0.4873046875,
74
+ 0.576171875
75
+ ],
76
+ "retrieval_batch_size": 32,
77
+ "retrieval_num_batches": 32,
78
+ "retrieval_num_samples": 1024,
79
+ "retrieval_order_policy": "manifest_order",
80
+ "retrieval_unique_datasets_per_batch": {
81
+ "max": 1.0,
82
+ "mean": 1.0,
83
+ "min": 1.0,
84
+ "std": 0.0
85
+ }
86
+ }
87
+ }
88
+ }
89
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/embedding_metrics.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "diversity_generated": 1.3626155066538168,
3
+ "embedding_frames": 60,
4
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
5
+ "matching_score_generated": 1.1581900119781494,
6
+ "matching_score_reference": 1.1037590503692627,
7
+ "motion_fid": 0.10886979363854854,
8
+ "motion_key": "body_pos_local",
9
+ "motion_kid": {
10
+ "mean": 6.802706844455386e-05,
11
+ "std": 0.0
12
+ },
13
+ "multimodality": null,
14
+ "num_samples": 1024,
15
+ "r_precision_generated": [
16
+ 0.5107421875,
17
+ 0.65625,
18
+ 0.7451171875
19
+ ],
20
+ "r_precision_reference": [
21
+ 0.6396484375,
22
+ 0.8076171875,
23
+ 0.8818359375
24
+ ],
25
+ "text_cache_dir": "${OMG_REPO_ROOT}/outputs/evaluator_text_cache/filtered_original_g1_tmr_bodypos_60f",
26
+ "text_encoder_model": "${OMG_REPO_ROOT}/models/t5-3b-local",
27
+ "text_generated": {
28
+ "matching_distance": {
29
+ "max": 1.5298819541931152,
30
+ "mean": 1.158189934794791,
31
+ "min": 0.8833644986152649,
32
+ "std": 0.1090334975205812
33
+ },
34
+ "matching_score": 1.1581900119781494,
35
+ "mean_text_rank": 3.1650390625,
36
+ "median_text_rank": 1.0,
37
+ "r_precision": [
38
+ 0.5107421875,
39
+ 0.65625,
40
+ 0.7451171875
41
+ ],
42
+ "retrieval_batch_size": 32,
43
+ "retrieval_num_batches": 32,
44
+ "retrieval_num_samples": 1024,
45
+ "retrieval_order_policy": "stratified_by_dataset",
46
+ "retrieval_unique_datasets_per_batch": {
47
+ "max": 1.0,
48
+ "mean": 1.0,
49
+ "min": 1.0,
50
+ "std": 0.0
51
+ }
52
+ },
53
+ "text_reference": {
54
+ "matching_distance": {
55
+ "max": 1.524587631225586,
56
+ "mean": 1.1037589554325677,
57
+ "min": 0.8309325575828552,
58
+ "std": 0.10401542255778194
59
+ },
60
+ "matching_score": 1.1037590503692627,
61
+ "mean_text_rank": 2.029296875,
62
+ "median_text_rank": 1.0,
63
+ "r_precision": [
64
+ 0.6396484375,
65
+ 0.8076171875,
66
+ 0.8818359375
67
+ ],
68
+ "retrieval_batch_size": 32,
69
+ "retrieval_num_batches": 32,
70
+ "retrieval_num_samples": 1024,
71
+ "retrieval_order_policy": "stratified_by_dataset",
72
+ "retrieval_unique_datasets_per_batch": {
73
+ "max": 1.0,
74
+ "mean": 1.0,
75
+ "min": 1.0,
76
+ "std": 0.0
77
+ }
78
+ },
79
+ "text_retrieval_batch_size": 32
80
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/metrics.json ADDED
@@ -0,0 +1,268 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark": {
3
+ "benchmark": "artifact_text",
4
+ "dataset_metrics": {
5
+ "filtered_original_mixed_modalities_all_materialized_test": {
6
+ "embedding": {
7
+ "diversity_generated": 1.3626155066538168,
8
+ "motion_fid": 0.10886979363854854,
9
+ "motion_kid": {
10
+ "mean": 6.802706844455386e-05,
11
+ "std": 0.0
12
+ },
13
+ "num_samples": 1024
14
+ },
15
+ "num_samples": 1024,
16
+ "physical": {
17
+ "body_jerk_mean": {
18
+ "max": 360.24395751953125,
19
+ "mean": 53.740967999212444,
20
+ "min": 1.9295663833618164,
21
+ "std": 55.003332467755946
22
+ },
23
+ "contact_sliding_speed": {
24
+ "max": 3.483940601348877,
25
+ "mean": 0.4634235018760364,
26
+ "min": 0.0,
27
+ "std": 0.573620541786601
28
+ },
29
+ "foot_ground_error": {
30
+ "max": 0.2830135226249695,
31
+ "mean": 0.04558111644996643,
32
+ "min": 0.001008824910968542,
33
+ "std": 0.029560935890020977
34
+ },
35
+ "num_samples": 1024
36
+ },
37
+ "text": {
38
+ "generated": {
39
+ "matching_distance": {
40
+ "max": 1.5298819541931152,
41
+ "mean": 1.158189934794791,
42
+ "min": 0.8833644986152649,
43
+ "std": 0.1090334975205812
44
+ },
45
+ "matching_score": 1.1581900119781494,
46
+ "mean_text_rank": 6.603515625,
47
+ "median_text_rank": 4.0,
48
+ "r_precision": [
49
+ 0.30078125,
50
+ 0.4052734375,
51
+ 0.4970703125
52
+ ],
53
+ "retrieval_batch_size": 32,
54
+ "retrieval_num_batches": 32,
55
+ "retrieval_num_samples": 1024,
56
+ "retrieval_order_policy": "manifest_order",
57
+ "retrieval_unique_datasets_per_batch": {
58
+ "max": 1.0,
59
+ "mean": 1.0,
60
+ "min": 1.0,
61
+ "std": 0.0
62
+ }
63
+ },
64
+ "reference": {
65
+ "matching_distance": {
66
+ "max": 1.524587631225586,
67
+ "mean": 1.1037589554325677,
68
+ "min": 0.8309325575828552,
69
+ "std": 0.10401542255778194
70
+ },
71
+ "matching_score": 1.1037590503692627,
72
+ "mean_text_rank": 5.2646484375,
73
+ "median_text_rank": 3.0,
74
+ "r_precision": [
75
+ 0.361328125,
76
+ 0.4873046875,
77
+ 0.576171875
78
+ ],
79
+ "retrieval_batch_size": 32,
80
+ "retrieval_num_batches": 32,
81
+ "retrieval_num_samples": 1024,
82
+ "retrieval_order_policy": "manifest_order",
83
+ "retrieval_unique_datasets_per_batch": {
84
+ "max": 1.0,
85
+ "mean": 1.0,
86
+ "min": 1.0,
87
+ "std": 0.0
88
+ }
89
+ }
90
+ }
91
+ }
92
+ },
93
+ "embedding_frames": 60,
94
+ "embedding_metrics": {
95
+ "diversity_generated": 1.3626155066538168,
96
+ "embedding_frames": 60,
97
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
98
+ "matching_score_generated": 1.1581900119781494,
99
+ "matching_score_reference": 1.1037590503692627,
100
+ "motion_fid": 0.10886979363854854,
101
+ "motion_key": "body_pos_local",
102
+ "motion_kid": {
103
+ "mean": 6.802706844455386e-05,
104
+ "std": 0.0
105
+ },
106
+ "multimodality": null,
107
+ "num_samples": 1024,
108
+ "r_precision_generated": [
109
+ 0.5107421875,
110
+ 0.65625,
111
+ 0.7451171875
112
+ ],
113
+ "r_precision_reference": [
114
+ 0.6396484375,
115
+ 0.8076171875,
116
+ 0.8818359375
117
+ ],
118
+ "text_cache_dir": "${OMG_REPO_ROOT}/outputs/evaluator_text_cache/filtered_original_g1_tmr_bodypos_60f",
119
+ "text_encoder_model": "${OMG_REPO_ROOT}/models/t5-3b-local",
120
+ "text_generated": {
121
+ "matching_distance": {
122
+ "max": 1.5298819541931152,
123
+ "mean": 1.158189934794791,
124
+ "min": 0.8833644986152649,
125
+ "std": 0.1090334975205812
126
+ },
127
+ "matching_score": 1.1581900119781494,
128
+ "mean_text_rank": 3.1650390625,
129
+ "median_text_rank": 1.0,
130
+ "r_precision": [
131
+ 0.5107421875,
132
+ 0.65625,
133
+ 0.7451171875
134
+ ],
135
+ "retrieval_batch_size": 32,
136
+ "retrieval_num_batches": 32,
137
+ "retrieval_num_samples": 1024,
138
+ "retrieval_order_policy": "stratified_by_dataset",
139
+ "retrieval_unique_datasets_per_batch": {
140
+ "max": 1.0,
141
+ "mean": 1.0,
142
+ "min": 1.0,
143
+ "std": 0.0
144
+ }
145
+ },
146
+ "text_reference": {
147
+ "matching_distance": {
148
+ "max": 1.524587631225586,
149
+ "mean": 1.1037589554325677,
150
+ "min": 0.8309325575828552,
151
+ "std": 0.10401542255778194
152
+ },
153
+ "matching_score": 1.1037590503692627,
154
+ "mean_text_rank": 2.029296875,
155
+ "median_text_rank": 1.0,
156
+ "r_precision": [
157
+ 0.6396484375,
158
+ 0.8076171875,
159
+ 0.8818359375
160
+ ],
161
+ "retrieval_batch_size": 32,
162
+ "retrieval_num_batches": 32,
163
+ "retrieval_num_samples": 1024,
164
+ "retrieval_order_policy": "stratified_by_dataset",
165
+ "retrieval_unique_datasets_per_batch": {
166
+ "max": 1.0,
167
+ "mean": 1.0,
168
+ "min": 1.0,
169
+ "std": 0.0
170
+ }
171
+ },
172
+ "text_retrieval_batch_size": 32
173
+ },
174
+ "exp": "g1_transformer_guided_300m",
175
+ "generated_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000/generated_qpos.npz",
176
+ "label": "OmniMotion 300M mixed-modal s045000",
177
+ "num_frames": 60,
178
+ "num_samples": 1024,
179
+ "physical_metrics": {
180
+ "body_jerk_mean": {
181
+ "max": 360.24395751953125,
182
+ "mean": 53.740967999212444,
183
+ "min": 1.9295663833618164,
184
+ "std": 55.003332467755946
185
+ },
186
+ "contact_height_threshold": 0.12,
187
+ "contact_penetration_tolerance": 0.02,
188
+ "contact_sliding_speed": {
189
+ "max": 3.483940601348877,
190
+ "mean": 0.4634235018760364,
191
+ "min": 0.0,
192
+ "std": 0.573620541786601
193
+ },
194
+ "foot_ground_error": {
195
+ "max": 0.2830135226249695,
196
+ "mean": 0.04558111644996643,
197
+ "min": 0.001008824910968542,
198
+ "std": 0.029560935890020977
199
+ },
200
+ "frames": 60,
201
+ "num_samples": 1024
202
+ },
203
+ "real_motion_metrics": {
204
+ "diversity_real": 1.3814387730607542,
205
+ "matching_distance_real": {
206
+ "max": 1.524587631225586,
207
+ "mean": 1.1037589554325677,
208
+ "min": 0.8309325575828552,
209
+ "std": 0.10401542255778194
210
+ },
211
+ "matching_score_real": 1.1037590503692627,
212
+ "motion_fid_reference_eval": 0.0,
213
+ "motion_fid_split_half": 0.0,
214
+ "motion_kid_reference_eval": {
215
+ "mean": -1.0798661567168466e-05,
216
+ "std": 0.0
217
+ },
218
+ "motion_kid_split_half": {
219
+ "mean": -1.0798661567168466e-05,
220
+ "std": 0.0
221
+ },
222
+ "multimodality": "-",
223
+ "num_reference_samples": 1024,
224
+ "num_samples": 1024,
225
+ "r_precision_real": [
226
+ 0.361328125,
227
+ 0.4873046875,
228
+ 0.576171875
229
+ ],
230
+ "text_real": {
231
+ "matching_distance": {
232
+ "max": 1.524587631225586,
233
+ "mean": 1.1037589554325677,
234
+ "min": 0.8309325575828552,
235
+ "std": 0.10401542255778194
236
+ },
237
+ "matching_score": 1.1037590503692627,
238
+ "mean_text_rank": 5.2646484375,
239
+ "median_text_rank": 3.0,
240
+ "r_precision": [
241
+ 0.361328125,
242
+ 0.4873046875,
243
+ 0.576171875
244
+ ],
245
+ "retrieval_batch_size": 32,
246
+ "retrieval_num_batches": 32,
247
+ "retrieval_num_samples": 1024,
248
+ "retrieval_order_policy": "manifest_order",
249
+ "retrieval_unique_datasets_per_batch": null
250
+ }
251
+ },
252
+ "reference_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/reference_qpos.npz",
253
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/sample_metrics.jsonl",
254
+ "sample_rankings_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/sample_rankings.json",
255
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000/samples.jsonl",
256
+ "split": "test",
257
+ "tracker_executed": {
258
+ "enabled": false
259
+ },
260
+ "transition_metrics": {
261
+ "chunk_length": 60,
262
+ "enabled": false,
263
+ "num_frames": 60,
264
+ "reason": "num_frames must be greater than chunk_length"
265
+ }
266
+ },
267
+ "label": "OmniMotion 300M mixed-modal s045000"
268
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/physical_metrics.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "body_jerk_mean": {
3
+ "max": 360.24395751953125,
4
+ "mean": 53.740967999212444,
5
+ "min": 1.9295663833618164,
6
+ "std": 55.003332467755946
7
+ },
8
+ "contact_height_threshold": 0.12,
9
+ "contact_penetration_tolerance": 0.02,
10
+ "contact_sliding_speed": {
11
+ "max": 3.483940601348877,
12
+ "mean": 0.4634235018760364,
13
+ "min": 0.0,
14
+ "std": 0.573620541786601
15
+ },
16
+ "foot_ground_error": {
17
+ "max": 0.2830135226249695,
18
+ "mean": 0.04558111644996643,
19
+ "min": 0.001008824910968542,
20
+ "std": 0.029560935890020977
21
+ },
22
+ "frames": 60,
23
+ "num_samples": 1024
24
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/real_motion_metrics.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "diversity_real": 1.3814387730607542,
3
+ "matching_distance_real": {
4
+ "max": 1.524587631225586,
5
+ "mean": 1.1037589554325677,
6
+ "min": 0.8309325575828552,
7
+ "std": 0.10401542255778194
8
+ },
9
+ "matching_score_real": 1.1037590503692627,
10
+ "motion_fid_reference_eval": 0.0,
11
+ "motion_fid_split_half": 0.0,
12
+ "motion_kid_reference_eval": {
13
+ "mean": -1.0798661567168466e-05,
14
+ "std": 0.0
15
+ },
16
+ "motion_kid_split_half": {
17
+ "mean": -1.0798661567168466e-05,
18
+ "std": 0.0
19
+ },
20
+ "multimodality": "-",
21
+ "num_reference_samples": 1024,
22
+ "num_samples": 1024,
23
+ "r_precision_real": [
24
+ 0.361328125,
25
+ 0.4873046875,
26
+ 0.576171875
27
+ ],
28
+ "text_real": {
29
+ "matching_distance": {
30
+ "max": 1.524587631225586,
31
+ "mean": 1.1037589554325677,
32
+ "min": 0.8309325575828552,
33
+ "std": 0.10401542255778194
34
+ },
35
+ "matching_score": 1.1037590503692627,
36
+ "mean_text_rank": 5.2646484375,
37
+ "median_text_rank": 3.0,
38
+ "r_precision": [
39
+ 0.361328125,
40
+ 0.4873046875,
41
+ 0.576171875
42
+ ],
43
+ "retrieval_batch_size": 32,
44
+ "retrieval_num_batches": 32,
45
+ "retrieval_num_samples": 1024,
46
+ "retrieval_order_policy": "manifest_order",
47
+ "retrieval_unique_datasets_per_batch": null
48
+ }
49
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/sample_metrics.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/sample_rankings.json ADDED
@@ -0,0 +1,1113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_generated": [
3
+ {
4
+ "caption": "sks person propels themselves forward with several hops.",
5
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
6
+ "dataset_index": 181646,
7
+ "generated_matching_distance": 0.8833644986152649,
8
+ "generated_reference_distance_gap": -0.06563186645507812,
9
+ "generated_text_rank": 1,
10
+ "reference_matching_distance": 0.948996365070343,
11
+ "reference_text_rank": 1,
12
+ "sample_index": 878
13
+ },
14
+ {
15
+ "caption": "A person grasps the chair with both hands and then uses their left hand to lift the phone.",
16
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
17
+ "dataset_index": 177349,
18
+ "generated_matching_distance": 0.8839854598045349,
19
+ "generated_reference_distance_gap": -0.08094435930252075,
20
+ "generated_text_rank": 1,
21
+ "reference_matching_distance": 0.9649298191070557,
22
+ "reference_text_rank": 1,
23
+ "sample_index": 722
24
+ },
25
+ {
26
+ "caption": "sks person jumps forward several times.",
27
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
28
+ "dataset_index": 181824,
29
+ "generated_matching_distance": 0.8950909376144409,
30
+ "generated_reference_distance_gap": -0.004495561122894287,
31
+ "generated_text_rank": 1,
32
+ "reference_matching_distance": 0.8995864987373352,
33
+ "reference_text_rank": 1,
34
+ "sample_index": 884
35
+ },
36
+ {
37
+ "caption": "A person uses their right hand to lift the phone and their left hand to set it down.",
38
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
39
+ "dataset_index": 177640,
40
+ "generated_matching_distance": 0.9016048312187195,
41
+ "generated_reference_distance_gap": -0.035506606101989746,
42
+ "generated_text_rank": 1,
43
+ "reference_matching_distance": 0.9371114373207092,
44
+ "reference_text_rank": 1,
45
+ "sample_index": 733
46
+ },
47
+ {
48
+ "caption": "A person uses their right hand to lift the phone and their left hand to set it down.",
49
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
50
+ "dataset_index": 177638,
51
+ "generated_matching_distance": 0.9151448011398315,
52
+ "generated_reference_distance_gap": 0.034645140171051025,
53
+ "generated_text_rank": 1,
54
+ "reference_matching_distance": 0.8804996609687805,
55
+ "reference_text_rank": 1,
56
+ "sample_index": 732
57
+ },
58
+ {
59
+ "caption": "The person casually stands on their right knee placing their left bent leg forward with their head and bent arms up. They lower their head with their arms and then stand up from the knee.",
60
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
61
+ "dataset_index": 194405,
62
+ "generated_matching_distance": 0.9217137098312378,
63
+ "generated_reference_distance_gap": -0.10609984397888184,
64
+ "generated_text_rank": 1,
65
+ "reference_matching_distance": 1.0278135538101196,
66
+ "reference_text_rank": 1,
67
+ "sample_index": 1023
68
+ },
69
+ {
70
+ "caption": "He walks forward, repeatedly taking steps while alternating between reaching into its pockets, retrieving items, extending its arm forward and downward, and returning to a neutral stance.; style: rhythmic",
71
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
72
+ "dataset_index": 48084,
73
+ "generated_matching_distance": 0.9271435737609863,
74
+ "generated_reference_distance_gap": -0.01666390895843506,
75
+ "generated_text_rank": 1,
76
+ "reference_matching_distance": 0.9438074827194214,
77
+ "reference_text_rank": 1,
78
+ "sample_index": 457
79
+ },
80
+ {
81
+ "caption": "start jogging forward",
82
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
83
+ "dataset_index": 26522,
84
+ "generated_matching_distance": 0.9299917817115784,
85
+ "generated_reference_distance_gap": -0.010540366172790527,
86
+ "generated_text_rank": 1,
87
+ "reference_matching_distance": 0.9405321478843689,
88
+ "reference_text_rank": 1,
89
+ "sample_index": 203
90
+ },
91
+ {
92
+ "caption": "The person stands indifferently with their feet shoulder-width apart. Then, they extend their right arm forward, closing their fingers as if beckoning someone towards them. Next, they raise their right arm, bent at the elbow, to their right ear, shaking their fingers near it, and then extend their right arm forward again, repeating the previous motion while turning their head to their right.",
93
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
94
+ "dataset_index": 186139,
95
+ "generated_matching_distance": 0.930182695388794,
96
+ "generated_reference_distance_gap": -0.037418365478515625,
97
+ "generated_text_rank": 1,
98
+ "reference_matching_distance": 0.9676010608673096,
99
+ "reference_text_rank": 1,
100
+ "sample_index": 961
101
+ },
102
+ {
103
+ "caption": "person coming to a stop after jogging forward",
104
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
105
+ "dataset_index": 27010,
106
+ "generated_matching_distance": 0.9309765696525574,
107
+ "generated_reference_distance_gap": -0.0133170485496521,
108
+ "generated_text_rank": 1,
109
+ "reference_matching_distance": 0.9442936182022095,
110
+ "reference_text_rank": 1,
111
+ "sample_index": 207
112
+ },
113
+ {
114
+ "caption": "A person uses their left hand to open the microwave and both hands to pick up the keyboard.",
115
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
116
+ "dataset_index": 177982,
117
+ "generated_matching_distance": 0.9348422884941101,
118
+ "generated_reference_distance_gap": -0.0011879801750183105,
119
+ "generated_text_rank": 1,
120
+ "reference_matching_distance": 0.9360302686691284,
121
+ "reference_text_rank": 1,
122
+ "sample_index": 762
123
+ },
124
+ {
125
+ "caption": "He stands still, starts with arms folded, extends them forward, raises them upward above its head, then brings its hands together in a continuous, coordinated sequence mimicking a rocket launch.; style: smooth",
126
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
127
+ "dataset_index": 48091,
128
+ "generated_matching_distance": 0.9358973503112793,
129
+ "generated_reference_distance_gap": -0.1061561107635498,
130
+ "generated_text_rank": 1,
131
+ "reference_matching_distance": 1.042053461074829,
132
+ "reference_text_rank": 1,
133
+ "sample_index": 459
134
+ },
135
+ {
136
+ "caption": "A person is sitting slightly swaying to the sides with their head slightly lowered, arms bent at the elbows, holding something in front in their left hand, fingering it, and their half-bent right hand at their side.",
137
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
138
+ "dataset_index": 185821,
139
+ "generated_matching_distance": 0.9371862411499023,
140
+ "generated_reference_distance_gap": 0.011291444301605225,
141
+ "generated_text_rank": 1,
142
+ "reference_matching_distance": 0.9258947968482971,
143
+ "reference_text_rank": 1,
144
+ "sample_index": 957
145
+ },
146
+ {
147
+ "caption": "In a series, sks person kicks towards the open space.",
148
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
149
+ "dataset_index": 182475,
150
+ "generated_matching_distance": 0.938368022441864,
151
+ "generated_reference_distance_gap": -0.016638755798339844,
152
+ "generated_text_rank": 1,
153
+ "reference_matching_distance": 0.9550067782402039,
154
+ "reference_text_rank": 1,
155
+ "sample_index": 926
156
+ },
157
+ {
158
+ "caption": "character sorts out medium big heavy object with two hands from medium behind and puts it away to the behind high",
159
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
160
+ "dataset_index": 30645,
161
+ "generated_matching_distance": 0.9417890906333923,
162
+ "generated_reference_distance_gap": 0.008968949317932129,
163
+ "generated_text_rank": 1,
164
+ "reference_matching_distance": 0.9328201413154602,
165
+ "reference_text_rank": 1,
166
+ "sample_index": 224
167
+ },
168
+ {
169
+ "caption": "He walks across a space with alternating leg steps, intermittently stopping to perform arm and shoulder elevation, elbow flexion and extension, wrist adjustment, and finger movements to simulate taking pictures.; style: controlled",
170
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
171
+ "dataset_index": 48584,
172
+ "generated_matching_distance": 0.9437264800071716,
173
+ "generated_reference_distance_gap": -0.05523192882537842,
174
+ "generated_text_rank": 1,
175
+ "reference_matching_distance": 0.99895840883255,
176
+ "reference_text_rank": 1,
177
+ "sample_index": 471
178
+ },
179
+ {
180
+ "caption": "sks person pushes off the ground and performs several forward jumps.",
181
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
182
+ "dataset_index": 181406,
183
+ "generated_matching_distance": 0.9481167793273926,
184
+ "generated_reference_distance_gap": 0.009528934955596924,
185
+ "generated_text_rank": 1,
186
+ "reference_matching_distance": 0.9385878443717957,
187
+ "reference_text_rank": 1,
188
+ "sample_index": 867
189
+ },
190
+ {
191
+ "caption": "The person stands still in a neutral pose, briefly twitching their right hand twice at the beginning. The person's head is slightly turned to the left.",
192
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
193
+ "dataset_index": 184201,
194
+ "generated_matching_distance": 0.9487717151641846,
195
+ "generated_reference_distance_gap": -0.017798542976379395,
196
+ "generated_text_rank": 1,
197
+ "reference_matching_distance": 0.966570258140564,
198
+ "reference_text_rank": 1,
199
+ "sample_index": 950
200
+ },
201
+ {
202
+ "caption": "A person places the cup down with their right hand and then sets their right hand on the stool.",
203
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
204
+ "dataset_index": 177804,
205
+ "generated_matching_distance": 0.9503446221351624,
206
+ "generated_reference_distance_gap": 0.013349950313568115,
207
+ "generated_text_rank": 1,
208
+ "reference_matching_distance": 0.9369946718215942,
209
+ "reference_text_rank": 1,
210
+ "sample_index": 745
211
+ },
212
+ {
213
+ "caption": "A man stands in place, repeatedly moves one arm and hand back and forth toward its mouth, while keeping the rest of its body relatively stationary.; style: controlled",
214
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
215
+ "dataset_index": 49376,
216
+ "generated_matching_distance": 0.9526408314704895,
217
+ "generated_reference_distance_gap": 0.01607424020767212,
218
+ "generated_text_rank": 1,
219
+ "reference_matching_distance": 0.9365665912628174,
220
+ "reference_text_rank": 1,
221
+ "sample_index": 499
222
+ }
223
+ ],
224
+ "best_reference": [
225
+ {
226
+ "caption": "sks person performs a low kick.",
227
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
228
+ "dataset_index": 181480,
229
+ "generated_matching_distance": 1.1953835487365723,
230
+ "generated_reference_distance_gap": 0.36445099115371704,
231
+ "generated_text_rank": 2,
232
+ "reference_matching_distance": 0.8309325575828552,
233
+ "reference_text_rank": 1,
234
+ "sample_index": 869
235
+ },
236
+ {
237
+ "caption": "sks person is bouncing ahead on one foot.",
238
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
239
+ "dataset_index": 181344,
240
+ "generated_matching_distance": 1.061639428138733,
241
+ "generated_reference_distance_gap": 0.200863778591156,
242
+ "generated_text_rank": 1,
243
+ "reference_matching_distance": 0.8607756495475769,
244
+ "reference_text_rank": 1,
245
+ "sample_index": 863
246
+ },
247
+ {
248
+ "caption": "Throwing multiple times, sks person propels an object through the air.",
249
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
250
+ "dataset_index": 181721,
251
+ "generated_matching_distance": 1.1049785614013672,
252
+ "generated_reference_distance_gap": 0.23957675695419312,
253
+ "generated_text_rank": 2,
254
+ "reference_matching_distance": 0.8654018044471741,
255
+ "reference_text_rank": 1,
256
+ "sample_index": 881
257
+ },
258
+ {
259
+ "caption": "sks person is bouncing ahead on one foot.",
260
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
261
+ "dataset_index": 182334,
262
+ "generated_matching_distance": 0.980375349521637,
263
+ "generated_reference_distance_gap": 0.10779011249542236,
264
+ "generated_text_rank": 1,
265
+ "reference_matching_distance": 0.8725852370262146,
266
+ "reference_text_rank": 1,
267
+ "sample_index": 916
268
+ },
269
+ {
270
+ "caption": "sks takes several forward jumps.",
271
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
272
+ "dataset_index": 181855,
273
+ "generated_matching_distance": 1.1005406379699707,
274
+ "generated_reference_distance_gap": 0.22222769260406494,
275
+ "generated_text_rank": 1,
276
+ "reference_matching_distance": 0.8783129453659058,
277
+ "reference_text_rank": 1,
278
+ "sample_index": 888
279
+ },
280
+ {
281
+ "caption": "sks person hops forward on one leg, repeating the motion several times.",
282
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
283
+ "dataset_index": 181147,
284
+ "generated_matching_distance": 1.0545120239257812,
285
+ "generated_reference_distance_gap": 0.17518460750579834,
286
+ "generated_text_rank": 2,
287
+ "reference_matching_distance": 0.8793274164199829,
288
+ "reference_text_rank": 1,
289
+ "sample_index": 859
290
+ },
291
+ {
292
+ "caption": "A person uses their right hand to lift the phone and their left hand to set it down.",
293
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
294
+ "dataset_index": 177638,
295
+ "generated_matching_distance": 0.9151448011398315,
296
+ "generated_reference_distance_gap": 0.034645140171051025,
297
+ "generated_text_rank": 1,
298
+ "reference_matching_distance": 0.8804996609687805,
299
+ "reference_text_rank": 1,
300
+ "sample_index": 732
301
+ },
302
+ {
303
+ "caption": "a person side steps to the left.",
304
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
305
+ "dataset_index": 44206,
306
+ "generated_matching_distance": 1.0676231384277344,
307
+ "generated_reference_distance_gap": 0.18495666980743408,
308
+ "generated_text_rank": 1,
309
+ "reference_matching_distance": 0.8826664686203003,
310
+ "reference_text_rank": 1,
311
+ "sample_index": 403
312
+ },
313
+ {
314
+ "caption": "a person jumps on the spot.",
315
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
316
+ "dataset_index": 42428,
317
+ "generated_matching_distance": 1.0838947296142578,
318
+ "generated_reference_distance_gap": 0.20035803318023682,
319
+ "generated_text_rank": 1,
320
+ "reference_matching_distance": 0.883536696434021,
321
+ "reference_text_rank": 1,
322
+ "sample_index": 382
323
+ },
324
+ {
325
+ "caption": "The person stands, wearily bringing their right hand to their mouth, leaning their torso forward, and nodding their head from side to side while lowering their right hand. They then raise both hands to their cheeks, shaking them near their face.",
326
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
327
+ "dataset_index": 194266,
328
+ "generated_matching_distance": 1.1928728818893433,
329
+ "generated_reference_distance_gap": 0.30654555559158325,
330
+ "generated_text_rank": 1,
331
+ "reference_matching_distance": 0.88632732629776,
332
+ "reference_text_rank": 1,
333
+ "sample_index": 1021
334
+ },
335
+ {
336
+ "caption": "A person is standing still with hands clasped in prayer, radiating calmness and serenity.",
337
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
338
+ "dataset_index": 109110,
339
+ "generated_matching_distance": 1.0945054292678833,
340
+ "generated_reference_distance_gap": 0.20755445957183838,
341
+ "generated_text_rank": 1,
342
+ "reference_matching_distance": 0.8869509696960449,
343
+ "reference_text_rank": 1,
344
+ "sample_index": 640
345
+ },
346
+ {
347
+ "caption": "sks punches straight ahead with one arm several times.",
348
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
349
+ "dataset_index": 182430,
350
+ "generated_matching_distance": 0.9872812032699585,
351
+ "generated_reference_distance_gap": 0.09422469139099121,
352
+ "generated_text_rank": 2,
353
+ "reference_matching_distance": 0.8930565118789673,
354
+ "reference_text_rank": 1,
355
+ "sample_index": 920
356
+ },
357
+ {
358
+ "caption": "sks person targets an object on the floor with several kicks.",
359
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
360
+ "dataset_index": 180917,
361
+ "generated_matching_distance": 1.2397817373275757,
362
+ "generated_reference_distance_gap": 0.3453025221824646,
363
+ "generated_text_rank": 1,
364
+ "reference_matching_distance": 0.8944792151451111,
365
+ "reference_text_rank": 1,
366
+ "sample_index": 855
367
+ },
368
+ {
369
+ "caption": "A man walks forward, alternating between legs with each step while flexing its knees to bring its heels toward its buttocks.; style: rhythmic",
370
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
371
+ "dataset_index": 48495,
372
+ "generated_matching_distance": 1.036427617073059,
373
+ "generated_reference_distance_gap": 0.1403970718383789,
374
+ "generated_text_rank": 1,
375
+ "reference_matching_distance": 0.8960305452346802,
376
+ "reference_text_rank": 1,
377
+ "sample_index": 468
378
+ },
379
+ {
380
+ "caption": "sks person hops forward on one leg, repeating the motion several times.",
381
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
382
+ "dataset_index": 180911,
383
+ "generated_matching_distance": 1.2365882396697998,
384
+ "generated_reference_distance_gap": 0.33717209100723267,
385
+ "generated_text_rank": 7,
386
+ "reference_matching_distance": 0.8994161486625671,
387
+ "reference_text_rank": 1,
388
+ "sample_index": 854
389
+ },
390
+ {
391
+ "caption": "sks person makes several throwing motions, sending an item forward.",
392
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
393
+ "dataset_index": 182488,
394
+ "generated_matching_distance": 0.9606974124908447,
395
+ "generated_reference_distance_gap": 0.06111687421798706,
396
+ "generated_text_rank": 1,
397
+ "reference_matching_distance": 0.8995805382728577,
398
+ "reference_text_rank": 1,
399
+ "sample_index": 927
400
+ },
401
+ {
402
+ "caption": "sks person jumps forward several times.",
403
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
404
+ "dataset_index": 181824,
405
+ "generated_matching_distance": 0.8950909376144409,
406
+ "generated_reference_distance_gap": -0.004495561122894287,
407
+ "generated_text_rank": 1,
408
+ "reference_matching_distance": 0.8995864987373352,
409
+ "reference_text_rank": 1,
410
+ "sample_index": 884
411
+ },
412
+ {
413
+ "caption": "Punching forward several times, sks person uses one arm.",
414
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
415
+ "dataset_index": 181868,
416
+ "generated_matching_distance": 0.9698982834815979,
417
+ "generated_reference_distance_gap": 0.06696736812591553,
418
+ "generated_text_rank": 2,
419
+ "reference_matching_distance": 0.9029309153556824,
420
+ "reference_text_rank": 1,
421
+ "sample_index": 889
422
+ },
423
+ {
424
+ "caption": "a person walks forward and then up stairs.",
425
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
426
+ "dataset_index": 41795,
427
+ "generated_matching_distance": 1.0723856687545776,
428
+ "generated_reference_distance_gap": 0.1670035719871521,
429
+ "generated_text_rank": 1,
430
+ "reference_matching_distance": 0.9053820967674255,
431
+ "reference_text_rank": 1,
432
+ "sample_index": 374
433
+ },
434
+ {
435
+ "caption": "sks person waves to someone, moving one arm repeatedly.",
436
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
437
+ "dataset_index": 181400,
438
+ "generated_matching_distance": 1.128285527229309,
439
+ "generated_reference_distance_gap": 0.21950608491897583,
440
+ "generated_text_rank": 1,
441
+ "reference_matching_distance": 0.9087794423103333,
442
+ "reference_text_rank": 1,
443
+ "sample_index": 866
444
+ }
445
+ ],
446
+ "largest_generated_reference_distance_gap": [
447
+ {
448
+ "caption": "a person jumps into the air.",
449
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
450
+ "dataset_index": 41468,
451
+ "generated_matching_distance": 1.4752897024154663,
452
+ "generated_reference_distance_gap": 0.4829879403114319,
453
+ "generated_text_rank": 30,
454
+ "reference_matching_distance": 0.9923017621040344,
455
+ "reference_text_rank": 1,
456
+ "sample_index": 367
457
+ },
458
+ {
459
+ "caption": "A person uses the left hand to pick up and put down the water bottle, and the right hand to open the fridge.",
460
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
461
+ "dataset_index": 177997,
462
+ "generated_matching_distance": 1.3580721616744995,
463
+ "generated_reference_distance_gap": 0.4413492679595947,
464
+ "generated_text_rank": 11,
465
+ "reference_matching_distance": 0.9167228937149048,
466
+ "reference_text_rank": 1,
467
+ "sample_index": 765
468
+ },
469
+ {
470
+ "caption": "A person is doing dumbbell bicep curls.",
471
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
472
+ "dataset_index": 176688,
473
+ "generated_matching_distance": 1.3814526796340942,
474
+ "generated_reference_distance_gap": 0.42044031620025635,
475
+ "generated_text_rank": 8,
476
+ "reference_matching_distance": 0.9610123634338379,
477
+ "reference_text_rank": 1,
478
+ "sample_index": 684
479
+ },
480
+ {
481
+ "caption": "A person is doing the Dumbbell Curl Trifecta exercise.",
482
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
483
+ "dataset_index": 176976,
484
+ "generated_matching_distance": 1.2659356594085693,
485
+ "generated_reference_distance_gap": 0.3843395709991455,
486
+ "generated_text_rank": 6,
487
+ "reference_matching_distance": 0.8815960884094238,
488
+ "reference_text_rank": 2,
489
+ "sample_index": 700
490
+ },
491
+ {
492
+ "caption": "stand in place with minimal leg and torso adjustments to maintain balance; style: stationary",
493
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
494
+ "dataset_index": 10905,
495
+ "generated_matching_distance": 1.4109145402908325,
496
+ "generated_reference_distance_gap": 0.371814489364624,
497
+ "generated_text_rank": 17,
498
+ "reference_matching_distance": 1.0391000509262085,
499
+ "reference_text_rank": 1,
500
+ "sample_index": 148
501
+ },
502
+ {
503
+ "caption": "character sorts out small heavy object with two hands from low up front and puts it away low behind",
504
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
505
+ "dataset_index": 34388,
506
+ "generated_matching_distance": 1.4242736101150513,
507
+ "generated_reference_distance_gap": 0.37024128437042236,
508
+ "generated_text_rank": 17,
509
+ "reference_matching_distance": 1.054032325744629,
510
+ "reference_text_rank": 1,
511
+ "sample_index": 240
512
+ },
513
+ {
514
+ "caption": "sks person performs a low kick.",
515
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
516
+ "dataset_index": 181480,
517
+ "generated_matching_distance": 1.1953835487365723,
518
+ "generated_reference_distance_gap": 0.36445099115371704,
519
+ "generated_text_rank": 2,
520
+ "reference_matching_distance": 0.8309325575828552,
521
+ "reference_text_rank": 1,
522
+ "sample_index": 869
523
+ },
524
+ {
525
+ "caption": "sks person targets an object on the floor with multiple kicks.",
526
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
527
+ "dataset_index": 182471,
528
+ "generated_matching_distance": 1.330911636352539,
529
+ "generated_reference_distance_gap": 0.35676074028015137,
530
+ "generated_text_rank": 9,
531
+ "reference_matching_distance": 0.9741508960723877,
532
+ "reference_text_rank": 1,
533
+ "sample_index": 925
534
+ },
535
+ {
536
+ "caption": "A woman performs a dynamic dance, starting with an upright stance and raised arms, transitioning through a series of poses and movements, including a squat, before returning to her initial position, exuding energy and confidence in her modern dance style.",
537
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
538
+ "dataset_index": 71465,
539
+ "generated_matching_distance": 1.5298819541931152,
540
+ "generated_reference_distance_gap": 0.3539113998413086,
541
+ "generated_text_rank": 29,
542
+ "reference_matching_distance": 1.1759705543518066,
543
+ "reference_text_rank": 1,
544
+ "sample_index": 613
545
+ },
546
+ {
547
+ "caption": "sks directs their kicks forward, making several motions.",
548
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
549
+ "dataset_index": 181786,
550
+ "generated_matching_distance": 1.2844429016113281,
551
+ "generated_reference_distance_gap": 0.35033005475997925,
552
+ "generated_text_rank": 6,
553
+ "reference_matching_distance": 0.9341128468513489,
554
+ "reference_text_rank": 1,
555
+ "sample_index": 882
556
+ },
557
+ {
558
+ "caption": "Sport fitness enthusiasts engage in left and right arm lifts while standing.",
559
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
560
+ "dataset_index": 39261,
561
+ "generated_matching_distance": 1.3618061542510986,
562
+ "generated_reference_distance_gap": 0.349798321723938,
563
+ "generated_text_rank": 11,
564
+ "reference_matching_distance": 1.0120078325271606,
565
+ "reference_text_rank": 1,
566
+ "sample_index": 289
567
+ },
568
+ {
569
+ "caption": "sks walks directly forward, following a straight path.",
570
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
571
+ "dataset_index": 182273,
572
+ "generated_matching_distance": 1.326947808265686,
573
+ "generated_reference_distance_gap": 0.34770095348358154,
574
+ "generated_text_rank": 19,
575
+ "reference_matching_distance": 0.9792468547821045,
576
+ "reference_text_rank": 1,
577
+ "sample_index": 913
578
+ },
579
+ {
580
+ "caption": "sks person targets an object on the floor with several kicks.",
581
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
582
+ "dataset_index": 180917,
583
+ "generated_matching_distance": 1.2397817373275757,
584
+ "generated_reference_distance_gap": 0.3453025221824646,
585
+ "generated_text_rank": 1,
586
+ "reference_matching_distance": 0.8944792151451111,
587
+ "reference_text_rank": 1,
588
+ "sample_index": 855
589
+ },
590
+ {
591
+ "caption": "person suddenly raises a finger in the air as if having an idea",
592
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
593
+ "dataset_index": 20930,
594
+ "generated_matching_distance": 1.2564125061035156,
595
+ "generated_reference_distance_gap": 0.3385847210884094,
596
+ "generated_text_rank": 5,
597
+ "reference_matching_distance": 0.9178277850151062,
598
+ "reference_text_rank": 1,
599
+ "sample_index": 194
600
+ },
601
+ {
602
+ "caption": "sks person hops forward on one leg, repeating the motion several times.",
603
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
604
+ "dataset_index": 180911,
605
+ "generated_matching_distance": 1.2365882396697998,
606
+ "generated_reference_distance_gap": 0.33717209100723267,
607
+ "generated_text_rank": 7,
608
+ "reference_matching_distance": 0.8994161486625671,
609
+ "reference_text_rank": 1,
610
+ "sample_index": 854
611
+ },
612
+ {
613
+ "caption": "sks person executes a series of kicks.",
614
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
615
+ "dataset_index": 182254,
616
+ "generated_matching_distance": 1.2027438879013062,
617
+ "generated_reference_distance_gap": 0.3250649571418762,
618
+ "generated_text_rank": 3,
619
+ "reference_matching_distance": 0.8776789307594299,
620
+ "reference_text_rank": 2,
621
+ "sample_index": 910
622
+ },
623
+ {
624
+ "caption": "A person closes the laptop with their right hand before picking it up with both hands.",
625
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
626
+ "dataset_index": 177298,
627
+ "generated_matching_distance": 1.2412724494934082,
628
+ "generated_reference_distance_gap": 0.32089847326278687,
629
+ "generated_text_rank": 2,
630
+ "reference_matching_distance": 0.9203739762306213,
631
+ "reference_text_rank": 1,
632
+ "sample_index": 718
633
+ },
634
+ {
635
+ "caption": "sks person performs a low kick.",
636
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
637
+ "dataset_index": 182512,
638
+ "generated_matching_distance": 1.2345249652862549,
639
+ "generated_reference_distance_gap": 0.31794679164886475,
640
+ "generated_text_rank": 4,
641
+ "reference_matching_distance": 0.9165781736373901,
642
+ "reference_text_rank": 1,
643
+ "sample_index": 932
644
+ },
645
+ {
646
+ "caption": "a figure stretches by twisting from side to side with arms outstretched then rolls their arms as well",
647
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
648
+ "dataset_index": 46058,
649
+ "generated_matching_distance": 1.2624322175979614,
650
+ "generated_reference_distance_gap": 0.31580525636672974,
651
+ "generated_text_rank": 5,
652
+ "reference_matching_distance": 0.9466269612312317,
653
+ "reference_text_rank": 1,
654
+ "sample_index": 419
655
+ },
656
+ {
657
+ "caption": "spin 180 degrees to face camera, raise right arm upward, bend torso forward, lift left leg high, bring both hands toward left knee; style: dynamic",
658
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
659
+ "dataset_index": 3657,
660
+ "generated_matching_distance": 1.4258980751037598,
661
+ "generated_reference_distance_gap": 0.3138573169708252,
662
+ "generated_text_rank": 21,
663
+ "reference_matching_distance": 1.1120407581329346,
664
+ "reference_text_rank": 1,
665
+ "sample_index": 86
666
+ }
667
+ ],
668
+ "num_samples": 1024,
669
+ "worst_generated": [
670
+ {
671
+ "caption": "a person jumps into the air.",
672
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
673
+ "dataset_index": 41468,
674
+ "generated_matching_distance": 1.4752897024154663,
675
+ "generated_reference_distance_gap": 0.4829879403114319,
676
+ "generated_text_rank": 30,
677
+ "reference_matching_distance": 0.9923017621040344,
678
+ "reference_text_rank": 1,
679
+ "sample_index": 367
680
+ },
681
+ {
682
+ "caption": "A woman performs a dynamic dance, starting with an upright stance and raised arms, transitioning through a series of poses and movements, including a squat, before returning to her initial position, exuding energy and confidence in her modern dance style.",
683
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
684
+ "dataset_index": 71465,
685
+ "generated_matching_distance": 1.5298819541931152,
686
+ "generated_reference_distance_gap": 0.3539113998413086,
687
+ "generated_text_rank": 29,
688
+ "reference_matching_distance": 1.1759705543518066,
689
+ "reference_text_rank": 1,
690
+ "sample_index": 613
691
+ },
692
+ {
693
+ "caption": "extend both arms forward with palms open, move them outward, and step backward with left leg; style: controlled",
694
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
695
+ "dataset_index": 3779,
696
+ "generated_matching_distance": 1.434065341949463,
697
+ "generated_reference_distance_gap": 0.21267783641815186,
698
+ "generated_text_rank": 29,
699
+ "reference_matching_distance": 1.221387505531311,
700
+ "reference_text_rank": 2,
701
+ "sample_index": 88
702
+ },
703
+ {
704
+ "caption": "lower left arm to body side, rotate torso back to neutral, return to initial bent knee stance; style: controlled",
705
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
706
+ "dataset_index": 11362,
707
+ "generated_matching_distance": 1.415978193283081,
708
+ "generated_reference_distance_gap": 0.2678908109664917,
709
+ "generated_text_rank": 29,
710
+ "reference_matching_distance": 1.1480873823165894,
711
+ "reference_text_rank": 1,
712
+ "sample_index": 152
713
+ },
714
+ {
715
+ "caption": "lower arms and return to neutral standing position; style: calm",
716
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
717
+ "dataset_index": 179310,
718
+ "generated_matching_distance": 1.4445661306381226,
719
+ "generated_reference_distance_gap": 0.15396511554718018,
720
+ "generated_text_rank": 28,
721
+ "reference_matching_distance": 1.2906010150909424,
722
+ "reference_text_rank": 8,
723
+ "sample_index": 812
724
+ },
725
+ {
726
+ "caption": "Happily, sport fitness enthusiasts perform a standing left and right arm stretch.",
727
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
728
+ "dataset_index": 39285,
729
+ "generated_matching_distance": 1.4244428873062134,
730
+ "generated_reference_distance_gap": 0.08158016204833984,
731
+ "generated_text_rank": 28,
732
+ "reference_matching_distance": 1.3428627252578735,
733
+ "reference_text_rank": 8,
734
+ "sample_index": 310
735
+ },
736
+ {
737
+ "caption": "rotate to face away, bend torso forward, then straighten up; style: fluid",
738
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
739
+ "dataset_index": 178987,
740
+ "generated_matching_distance": 1.4176899194717407,
741
+ "generated_reference_distance_gap": 0.1974332332611084,
742
+ "generated_text_rank": 26,
743
+ "reference_matching_distance": 1.2202566862106323,
744
+ "reference_text_rank": 3,
745
+ "sample_index": 800
746
+ },
747
+ {
748
+ "caption": "the person is turning on the shower with her right arm.",
749
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
750
+ "dataset_index": 46506,
751
+ "generated_matching_distance": 1.462831974029541,
752
+ "generated_reference_distance_gap": 0.3073045015335083,
753
+ "generated_text_rank": 25,
754
+ "reference_matching_distance": 1.1555274724960327,
755
+ "reference_text_rank": 2,
756
+ "sample_index": 424
757
+ },
758
+ {
759
+ "caption": "He starts in a relaxed standing stance, leans back as if resting on an unseen support, then moves forward to return to an upright standing position.; style: smooth",
760
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
761
+ "dataset_index": 49364,
762
+ "generated_matching_distance": 1.4814107418060303,
763
+ "generated_reference_distance_gap": 0.2973593473434448,
764
+ "generated_text_rank": 24,
765
+ "reference_matching_distance": 1.1840513944625854,
766
+ "reference_text_rank": 3,
767
+ "sample_index": 497
768
+ },
769
+ {
770
+ "caption": "bend torso forward and extend arms downward; style: controlled",
771
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
772
+ "dataset_index": 180016,
773
+ "generated_matching_distance": 1.4214866161346436,
774
+ "generated_reference_distance_gap": 0.19545376300811768,
775
+ "generated_text_rank": 23,
776
+ "reference_matching_distance": 1.2260328531265259,
777
+ "reference_text_rank": 4,
778
+ "sample_index": 831
779
+ },
780
+ {
781
+ "caption": "spin 180 degrees to face camera, raise right arm upward, bend torso forward, lift left leg high, bring both hands toward left knee; style: dynamic",
782
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
783
+ "dataset_index": 3657,
784
+ "generated_matching_distance": 1.4258980751037598,
785
+ "generated_reference_distance_gap": 0.3138573169708252,
786
+ "generated_text_rank": 21,
787
+ "reference_matching_distance": 1.1120407581329346,
788
+ "reference_text_rank": 1,
789
+ "sample_index": 86
790
+ },
791
+ {
792
+ "caption": "stand still, then bend forward with hands on knees; style: calm",
793
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
794
+ "dataset_index": 50399,
795
+ "generated_matching_distance": 1.379555106163025,
796
+ "generated_reference_distance_gap": -0.08733487129211426,
797
+ "generated_text_rank": 21,
798
+ "reference_matching_distance": 1.4668899774551392,
799
+ "reference_text_rank": 21,
800
+ "sample_index": 580
801
+ },
802
+ {
803
+ "caption": "A woman is doing a stretching exercise with her feet shoulder-width apart and her arms extended towards a chair, keeping her back straight.",
804
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
805
+ "dataset_index": 144093,
806
+ "generated_matching_distance": 1.4389883279800415,
807
+ "generated_reference_distance_gap": 0.2030489444732666,
808
+ "generated_text_rank": 20,
809
+ "reference_matching_distance": 1.235939383506775,
810
+ "reference_text_rank": 1,
811
+ "sample_index": 665
812
+ },
813
+ {
814
+ "caption": "characterize a slow diagonal retreat",
815
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
816
+ "dataset_index": 27329,
817
+ "generated_matching_distance": 1.35233473777771,
818
+ "generated_reference_distance_gap": 0.2180042266845703,
819
+ "generated_text_rank": 20,
820
+ "reference_matching_distance": 1.1343305110931396,
821
+ "reference_text_rank": 3,
822
+ "sample_index": 208
823
+ },
824
+ {
825
+ "caption": "sks walks directly forward, following a straight path.",
826
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
827
+ "dataset_index": 182273,
828
+ "generated_matching_distance": 1.326947808265686,
829
+ "generated_reference_distance_gap": 0.34770095348358154,
830
+ "generated_text_rank": 19,
831
+ "reference_matching_distance": 0.9792468547821045,
832
+ "reference_text_rank": 1,
833
+ "sample_index": 913
834
+ },
835
+ {
836
+ "caption": "character sorts out small heavy object with two hands from low up front and puts it away low behind",
837
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
838
+ "dataset_index": 34388,
839
+ "generated_matching_distance": 1.4242736101150513,
840
+ "generated_reference_distance_gap": 0.37024128437042236,
841
+ "generated_text_rank": 17,
842
+ "reference_matching_distance": 1.054032325744629,
843
+ "reference_text_rank": 1,
844
+ "sample_index": 240
845
+ },
846
+ {
847
+ "caption": "stand in place with minimal leg and torso adjustments to maintain balance; style: stationary",
848
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
849
+ "dataset_index": 10905,
850
+ "generated_matching_distance": 1.4109145402908325,
851
+ "generated_reference_distance_gap": 0.371814489364624,
852
+ "generated_text_rank": 17,
853
+ "reference_matching_distance": 1.0391000509262085,
854
+ "reference_text_rank": 1,
855
+ "sample_index": 148
856
+ },
857
+ {
858
+ "caption": "Happily, sport fitness enthusiasts perform a standing left and right arm stretch.",
859
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
860
+ "dataset_index": 39292,
861
+ "generated_matching_distance": 1.3998994827270508,
862
+ "generated_reference_distance_gap": 0.0434802770614624,
863
+ "generated_text_rank": 17,
864
+ "reference_matching_distance": 1.3564192056655884,
865
+ "reference_text_rank": 7,
866
+ "sample_index": 315
867
+ },
868
+ {
869
+ "caption": "She starts by lifting her right leg and bending it at the knee, extending her left arm and curving her right arm.",
870
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
871
+ "dataset_index": 90516,
872
+ "generated_matching_distance": 1.3796504735946655,
873
+ "generated_reference_distance_gap": 0.10803878307342529,
874
+ "generated_text_rank": 17,
875
+ "reference_matching_distance": 1.2716116905212402,
876
+ "reference_text_rank": 4,
877
+ "sample_index": 627
878
+ },
879
+ {
880
+ "caption": "She performs ballet with a strong sense of revulsion.",
881
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
882
+ "dataset_index": 39226,
883
+ "generated_matching_distance": 1.322020411491394,
884
+ "generated_reference_distance_gap": 0.027699947357177734,
885
+ "generated_text_rank": 17,
886
+ "reference_matching_distance": 1.2943204641342163,
887
+ "reference_text_rank": 4,
888
+ "sample_index": 263
889
+ }
890
+ ],
891
+ "worst_reference": [
892
+ {
893
+ "caption": "step into wide bent-knee stance, hold position, then return to narrow stance; style: dynamic",
894
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
895
+ "dataset_index": 12170,
896
+ "generated_matching_distance": 1.3963404893875122,
897
+ "generated_reference_distance_gap": -0.12824714183807373,
898
+ "generated_text_rank": 9,
899
+ "reference_matching_distance": 1.524587631225586,
900
+ "reference_text_rank": 28,
901
+ "sample_index": 166
902
+ },
903
+ {
904
+ "caption": "stand still, then bend forward with hands on knees; style: calm",
905
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
906
+ "dataset_index": 50399,
907
+ "generated_matching_distance": 1.379555106163025,
908
+ "generated_reference_distance_gap": -0.08733487129211426,
909
+ "generated_text_rank": 21,
910
+ "reference_matching_distance": 1.4668899774551392,
911
+ "reference_text_rank": 21,
912
+ "sample_index": 580
913
+ },
914
+ {
915
+ "caption": "walk forward with alternating leg movements and gentle arm swings; style: rhythmic, controlled",
916
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
917
+ "dataset_index": 4549,
918
+ "generated_matching_distance": 1.1645960807800293,
919
+ "generated_reference_distance_gap": -0.14809024333953857,
920
+ "generated_text_rank": 2,
921
+ "reference_matching_distance": 1.3126863241195679,
922
+ "reference_text_rank": 20,
923
+ "sample_index": 94
924
+ },
925
+ {
926
+ "caption": "The humanoid robot performs an in-place dance routine, shifting weight between feet, slightly bending knees, and moving arms in coordination with leg motions.; style: rhythmic",
927
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
928
+ "dataset_index": 47131,
929
+ "generated_matching_distance": 1.1655032634735107,
930
+ "generated_reference_distance_gap": -0.23436427116394043,
931
+ "generated_text_rank": 1,
932
+ "reference_matching_distance": 1.3998675346374512,
933
+ "reference_text_rank": 16,
934
+ "sample_index": 433
935
+ },
936
+ {
937
+ "caption": "A man in athletic wear is sprinting forward, mimicking a marathon runner's powerful stride.",
938
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
939
+ "dataset_index": 114442,
940
+ "generated_matching_distance": 1.2486003637313843,
941
+ "generated_reference_distance_gap": -0.11720526218414307,
942
+ "generated_text_rank": 4,
943
+ "reference_matching_distance": 1.3658056259155273,
944
+ "reference_text_rank": 16,
945
+ "sample_index": 642
946
+ },
947
+ {
948
+ "caption": "rise from crouch to standing, extend both arms outward, and lift body slightly upward; style: smooth",
949
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
950
+ "dataset_index": 9127,
951
+ "generated_matching_distance": 1.280436635017395,
952
+ "generated_reference_distance_gap": -0.07181048393249512,
953
+ "generated_text_rank": 10,
954
+ "reference_matching_distance": 1.3522471189498901,
955
+ "reference_text_rank": 16,
956
+ "sample_index": 132
957
+ },
958
+ {
959
+ "caption": "alternate leg lifts with quick arm movements; style: energetic",
960
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
961
+ "dataset_index": 1869,
962
+ "generated_matching_distance": 1.1801400184631348,
963
+ "generated_reference_distance_gap": -0.19436395168304443,
964
+ "generated_text_rank": 5,
965
+ "reference_matching_distance": 1.3745039701461792,
966
+ "reference_text_rank": 14,
967
+ "sample_index": 40
968
+ },
969
+ {
970
+ "caption": "perform strutting dance with steps, turns, arm swings; style: rhythmic upbeat",
971
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
972
+ "dataset_index": 2813,
973
+ "generated_matching_distance": 1.2969133853912354,
974
+ "generated_reference_distance_gap": -0.011563658714294434,
975
+ "generated_text_rank": 12,
976
+ "reference_matching_distance": 1.3084770441055298,
977
+ "reference_text_rank": 14,
978
+ "sample_index": 68
979
+ },
980
+ {
981
+ "caption": "walk forward with coordinated arm and leg movements; style: steady",
982
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
983
+ "dataset_index": 180285,
984
+ "generated_matching_distance": 1.170807957649231,
985
+ "generated_reference_distance_gap": -0.13276910781860352,
986
+ "generated_text_rank": 6,
987
+ "reference_matching_distance": 1.3035770654678345,
988
+ "reference_text_rank": 14,
989
+ "sample_index": 837
990
+ },
991
+ {
992
+ "caption": "stand still, bend forward holding pose; style: static",
993
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
994
+ "dataset_index": 50432,
995
+ "generated_matching_distance": 1.2234742641448975,
996
+ "generated_reference_distance_gap": -0.05692183971405029,
997
+ "generated_text_rank": 2,
998
+ "reference_matching_distance": 1.2803961038589478,
999
+ "reference_text_rank": 14,
1000
+ "sample_index": 588
1001
+ },
1002
+ {
1003
+ "caption": "stand upright with arms extended horizontally; style: static",
1004
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1005
+ "dataset_index": 50451,
1006
+ "generated_matching_distance": 1.3834019899368286,
1007
+ "generated_reference_distance_gap": 0.022049903869628906,
1008
+ "generated_text_rank": 15,
1009
+ "reference_matching_distance": 1.3613520860671997,
1010
+ "reference_text_rank": 13,
1011
+ "sample_index": 589
1012
+ },
1013
+ {
1014
+ "caption": "dance with hip sways and alternating arm swings; style: energetic",
1015
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1016
+ "dataset_index": 50480,
1017
+ "generated_matching_distance": 1.2250499725341797,
1018
+ "generated_reference_distance_gap": -0.042949676513671875,
1019
+ "generated_text_rank": 7,
1020
+ "reference_matching_distance": 1.2679996490478516,
1021
+ "reference_text_rank": 13,
1022
+ "sample_index": 590
1023
+ },
1024
+ {
1025
+ "caption": "Runners sprint back and forth, fully focused.",
1026
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1027
+ "dataset_index": 39247,
1028
+ "generated_matching_distance": 1.2156513929367065,
1029
+ "generated_reference_distance_gap": -0.09586632251739502,
1030
+ "generated_text_rank": 2,
1031
+ "reference_matching_distance": 1.3115177154541016,
1032
+ "reference_text_rank": 12,
1033
+ "sample_index": 278
1034
+ },
1035
+ {
1036
+ "caption": "stand still facing backward; style: static",
1037
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1038
+ "dataset_index": 11469,
1039
+ "generated_matching_distance": 1.2919294834136963,
1040
+ "generated_reference_distance_gap": 0.00890660285949707,
1041
+ "generated_text_rank": 11,
1042
+ "reference_matching_distance": 1.2830228805541992,
1043
+ "reference_text_rank": 12,
1044
+ "sample_index": 153
1045
+ },
1046
+ {
1047
+ "caption": "dance with hip sways and alternating arm swings; style: energetic",
1048
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1049
+ "dataset_index": 50484,
1050
+ "generated_matching_distance": 1.333507776260376,
1051
+ "generated_reference_distance_gap": 0.08620452880859375,
1052
+ "generated_text_rank": 14,
1053
+ "reference_matching_distance": 1.2473032474517822,
1054
+ "reference_text_rank": 12,
1055
+ "sample_index": 591
1056
+ },
1057
+ {
1058
+ "caption": "Runners sprint back and forth, fully focused.",
1059
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1060
+ "dataset_index": 39246,
1061
+ "generated_matching_distance": 1.360196590423584,
1062
+ "generated_reference_distance_gap": 0.05237269401550293,
1063
+ "generated_text_rank": 14,
1064
+ "reference_matching_distance": 1.307823896408081,
1065
+ "reference_text_rank": 11,
1066
+ "sample_index": 277
1067
+ },
1068
+ {
1069
+ "caption": "execute dance with arm swings, steps, spins, and turns; style: rhythmic",
1070
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1071
+ "dataset_index": 1259,
1072
+ "generated_matching_distance": 1.2254981994628906,
1073
+ "generated_reference_distance_gap": -0.06908321380615234,
1074
+ "generated_text_rank": 5,
1075
+ "reference_matching_distance": 1.294581413269043,
1076
+ "reference_text_rank": 11,
1077
+ "sample_index": 24
1078
+ },
1079
+ {
1080
+ "caption": "stand still, then bend forward holding pose; style: static",
1081
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1082
+ "dataset_index": 50405,
1083
+ "generated_matching_distance": 1.324897289276123,
1084
+ "generated_reference_distance_gap": -0.09820139408111572,
1085
+ "generated_text_rank": 5,
1086
+ "reference_matching_distance": 1.4230986833572388,
1087
+ "reference_text_rank": 10,
1088
+ "sample_index": 582
1089
+ },
1090
+ {
1091
+ "caption": "a person leaning ove to fix his pant leg with the right hand and then readjusts his pant leg again with the left hand.",
1092
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1093
+ "dataset_index": 43924,
1094
+ "generated_matching_distance": 1.3439409732818604,
1095
+ "generated_reference_distance_gap": 0.02637648582458496,
1096
+ "generated_text_rank": 10,
1097
+ "reference_matching_distance": 1.3175644874572754,
1098
+ "reference_text_rank": 10,
1099
+ "sample_index": 398
1100
+ },
1101
+ {
1102
+ "caption": "The humanoid robot moves forward with one leg, bends knees slightly while halting, raises one arm forward in a stopping gesture, returns the arm to its side, then assumes a stable standing position with feet shoulder-width apart.; style: controlled",
1103
+ "dataset": "filtered_original_mixed_modalities_all_materialized_test",
1104
+ "dataset_index": 48697,
1105
+ "generated_matching_distance": 1.3533991575241089,
1106
+ "generated_reference_distance_gap": 0.004469752311706543,
1107
+ "generated_text_rank": 11,
1108
+ "reference_matching_distance": 1.3489294052124023,
1109
+ "reference_text_rank": 9,
1110
+ "sample_index": 472
1111
+ }
1112
+ ]
1113
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/samples.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/summary.md ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ # Generation Benchmark Summary
2
+
3
+ - benchmark: artifact_text
4
+ - label: OmniMotion 300M mixed-modal s045000
5
+ - exp: g1_transformer_guided_300m
6
+ - split: test
7
+
8
+ | ckpt | FID(seq) | KID(seq) | R@1 | R@2 | R@3 | Diversity | Matching | MultiModal Dist | BeatAlign | BeatAlign Ref | BeatAlign Gap | FIDk | FIDg | Divk | Divg | PFC | PFC Ref | contact_sliding | foot_ground_error | body_jerk | PJ | AUJ | Tracker g-MPJPE | Tracker MPJPE | Tracker E_vel | Tracker E_acc |
9
+ |---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
10
+ | OmniMotion 300M mixed-modal s045000 | **0.1089** | **0.0001 +/- 0.0000** | **0.5107** | **0.6562** | **0.7451** | **1.3626** | **1.1582** | **1.1582 +/- 0.1090** | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | **0.4634 +/- 0.5736** | **0.0456 +/- 0.0296** | **53.7410 +/- 55.0033** | N/A | N/A | N/A | N/A | N/A | N/A |
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/tracker_executed.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "enabled": false
3
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_s045000_textmetrics_artifact/transition_metrics.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 60,
3
+ "enabled": false,
4
+ "num_frames": 60,
5
+ "reason": "num_frames must be greater than chunk_length"
6
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/artifact_samples.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/benchmark.json ADDED
@@ -0,0 +1,2295 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark": "artifact_text",
3
+ "dataset_metrics": {
4
+ "100style_original_filtered_test": {
5
+ "embedding": {
6
+ "diversity_generated": 1.0575839117993924,
7
+ "motion_fid": 0.295132260782035,
8
+ "motion_kid": {
9
+ "mean": 0.00030837465139699205,
10
+ "std": 0.0
11
+ },
12
+ "num_samples": 86
13
+ },
14
+ "num_samples": 86,
15
+ "physical": {
16
+ "body_jerk_mean": {
17
+ "max": 306.9395446777344,
18
+ "mean": 91.73821926116943,
19
+ "min": 14.676246643066406,
20
+ "std": 41.61339123942521
21
+ },
22
+ "contact_sliding_speed": {
23
+ "max": 1.9230101108551025,
24
+ "mean": 0.7933437027318706,
25
+ "min": 0.0035282010212540627,
26
+ "std": 0.39022621340146474
27
+ },
28
+ "foot_ground_error": {
29
+ "max": 0.04590573161840439,
30
+ "mean": 0.013192172581871409,
31
+ "min": 0.0021116272546350956,
32
+ "std": 0.009524469604494283
33
+ },
34
+ "num_samples": 86
35
+ },
36
+ "text": {
37
+ "generated": {
38
+ "matching_distance": {
39
+ "max": 1.3606981039047241,
40
+ "mean": 1.1547354348870211,
41
+ "min": 0.9792652130126953,
42
+ "std": 0.07691267359783768
43
+ },
44
+ "matching_score": 1.1547354459762573,
45
+ "mean_text_rank": 7.109375,
46
+ "median_text_rank": 4.0,
47
+ "r_precision": [
48
+ 0.3125,
49
+ 0.421875,
50
+ 0.453125
51
+ ],
52
+ "retrieval_batch_size": 32,
53
+ "retrieval_num_batches": 2,
54
+ "retrieval_num_samples": 64,
55
+ "retrieval_order_policy": "manifest_order",
56
+ "retrieval_unique_datasets_per_batch": {
57
+ "max": 1.0,
58
+ "mean": 1.0,
59
+ "min": 1.0,
60
+ "std": 0.0
61
+ }
62
+ },
63
+ "reference": {
64
+ "matching_distance": {
65
+ "max": 1.3797343969345093,
66
+ "mean": 1.1303709336491519,
67
+ "min": 0.976786196231842,
68
+ "std": 0.0715656855131584
69
+ },
70
+ "matching_score": 1.130370855331421,
71
+ "mean_text_rank": 6.03125,
72
+ "median_text_rank": 4.0,
73
+ "r_precision": [
74
+ 0.328125,
75
+ 0.453125,
76
+ 0.46875
77
+ ],
78
+ "retrieval_batch_size": 32,
79
+ "retrieval_num_batches": 2,
80
+ "retrieval_num_samples": 64,
81
+ "retrieval_order_policy": "manifest_order",
82
+ "retrieval_unique_datasets_per_batch": {
83
+ "max": 1.0,
84
+ "mean": 1.0,
85
+ "min": 1.0,
86
+ "std": 0.0
87
+ }
88
+ }
89
+ }
90
+ },
91
+ "amass_original_filtered_test": {
92
+ "embedding": {
93
+ "diversity_generated": 1.2720565270928372,
94
+ "motion_fid": 0.3890069629224415,
95
+ "motion_kid": {
96
+ "mean": 7.148098609599529e-05,
97
+ "std": 0.0
98
+ },
99
+ "num_samples": 86
100
+ },
101
+ "num_samples": 86,
102
+ "physical": {
103
+ "body_jerk_mean": {
104
+ "max": 265.1717834472656,
105
+ "mean": 72.45789514585982,
106
+ "min": 11.791740417480469,
107
+ "std": 45.92659264260362
108
+ },
109
+ "contact_sliding_speed": {
110
+ "max": 1.6666642427444458,
111
+ "mean": 0.40420085963045876,
112
+ "min": 0.0,
113
+ "std": 0.4343716464602378
114
+ },
115
+ "foot_ground_error": {
116
+ "max": 0.11975693702697754,
117
+ "mean": 0.026942830335153916,
118
+ "min": 0.0033850527834147215,
119
+ "std": 0.016949850606657043
120
+ },
121
+ "num_samples": 86
122
+ },
123
+ "text": {
124
+ "generated": {
125
+ "matching_distance": {
126
+ "max": 1.5543382167816162,
127
+ "mean": 1.1741622436878294,
128
+ "min": 1.0039037466049194,
129
+ "std": 0.10332746152566549
130
+ },
131
+ "matching_score": 1.1741621494293213,
132
+ "mean_text_rank": 4.84375,
133
+ "median_text_rank": 3.0,
134
+ "r_precision": [
135
+ 0.265625,
136
+ 0.46875,
137
+ 0.5625
138
+ ],
139
+ "retrieval_batch_size": 32,
140
+ "retrieval_num_batches": 2,
141
+ "retrieval_num_samples": 64,
142
+ "retrieval_order_policy": "manifest_order",
143
+ "retrieval_unique_datasets_per_batch": {
144
+ "max": 1.0,
145
+ "mean": 1.0,
146
+ "min": 1.0,
147
+ "std": 0.0
148
+ }
149
+ },
150
+ "reference": {
151
+ "matching_distance": {
152
+ "max": 1.4848475456237793,
153
+ "mean": 1.1559161391369133,
154
+ "min": 0.9887187480926514,
155
+ "std": 0.09615433157926441
156
+ },
157
+ "matching_score": 1.1559160947799683,
158
+ "mean_text_rank": 4.71875,
159
+ "median_text_rank": 3.0,
160
+ "r_precision": [
161
+ 0.3125,
162
+ 0.4375,
163
+ 0.5625
164
+ ],
165
+ "retrieval_batch_size": 32,
166
+ "retrieval_num_batches": 2,
167
+ "retrieval_num_samples": 64,
168
+ "retrieval_order_policy": "manifest_order",
169
+ "retrieval_unique_datasets_per_batch": {
170
+ "max": 1.0,
171
+ "mean": 1.0,
172
+ "min": 1.0,
173
+ "std": 0.0
174
+ }
175
+ }
176
+ }
177
+ },
178
+ "bones_seed_original_filtered_test": {
179
+ "embedding": {
180
+ "diversity_generated": 1.2288775777117131,
181
+ "motion_fid": 0.2935687819807675,
182
+ "motion_kid": {
183
+ "mean": 4.6969598594959194e-05,
184
+ "std": 0.0
185
+ },
186
+ "num_samples": 86
187
+ },
188
+ "num_samples": 86,
189
+ "physical": {
190
+ "body_jerk_mean": {
191
+ "max": 392.7170104980469,
192
+ "mean": 67.97546840268512,
193
+ "min": 14.550009727478027,
194
+ "std": 66.54340159249884
195
+ },
196
+ "contact_sliding_speed": {
197
+ "max": 1.5325274467468262,
198
+ "mean": 0.32937751099099083,
199
+ "min": 0.004231189843267202,
200
+ "std": 0.4110081512966714
201
+ },
202
+ "foot_ground_error": {
203
+ "max": 0.06508437544107437,
204
+ "mean": 0.013941306205260632,
205
+ "min": 0.001276918570511043,
206
+ "std": 0.011664757008804184
207
+ },
208
+ "num_samples": 86
209
+ },
210
+ "text": {
211
+ "generated": {
212
+ "matching_distance": {
213
+ "max": 1.3572005033493042,
214
+ "mean": 1.1043026724527047,
215
+ "min": 0.8937378525733948,
216
+ "std": 0.11051084308488551
217
+ },
218
+ "matching_score": 1.1043027639389038,
219
+ "mean_text_rank": 5.359375,
220
+ "median_text_rank": 2.0,
221
+ "r_precision": [
222
+ 0.46875,
223
+ 0.578125,
224
+ 0.65625
225
+ ],
226
+ "retrieval_batch_size": 32,
227
+ "retrieval_num_batches": 2,
228
+ "retrieval_num_samples": 64,
229
+ "retrieval_order_policy": "manifest_order",
230
+ "retrieval_unique_datasets_per_batch": {
231
+ "max": 1.0,
232
+ "mean": 1.0,
233
+ "min": 1.0,
234
+ "std": 0.0
235
+ }
236
+ },
237
+ "reference": {
238
+ "matching_distance": {
239
+ "max": 1.2490898370742798,
240
+ "mean": 1.0238502039465793,
241
+ "min": 0.8655920624732971,
242
+ "std": 0.08327281471573043
243
+ },
244
+ "matching_score": 1.0238500833511353,
245
+ "mean_text_rank": 2.03125,
246
+ "median_text_rank": 1.0,
247
+ "r_precision": [
248
+ 0.734375,
249
+ 0.828125,
250
+ 0.84375
251
+ ],
252
+ "retrieval_batch_size": 32,
253
+ "retrieval_num_batches": 2,
254
+ "retrieval_num_samples": 64,
255
+ "retrieval_order_policy": "manifest_order",
256
+ "retrieval_unique_datasets_per_batch": {
257
+ "max": 1.0,
258
+ "mean": 1.0,
259
+ "min": 1.0,
260
+ "std": 0.0
261
+ }
262
+ }
263
+ }
264
+ },
265
+ "fitness_original_filtered_test": {
266
+ "embedding": {
267
+ "diversity_generated": 1.2788421887176629,
268
+ "motion_fid": 0.7282577019576995,
269
+ "motion_kid": {
270
+ "mean": 0.0006897810534862714,
271
+ "std": 0.0
272
+ },
273
+ "num_samples": 86
274
+ },
275
+ "num_samples": 86,
276
+ "physical": {
277
+ "body_jerk_mean": {
278
+ "max": 229.07168579101562,
279
+ "mean": 80.73191664939704,
280
+ "min": 22.525007247924805,
281
+ "std": 47.517861751693125
282
+ },
283
+ "contact_sliding_speed": {
284
+ "max": 1.1983479261398315,
285
+ "mean": 0.1302262063917899,
286
+ "min": 0.0,
287
+ "std": 0.19028626905031584
288
+ },
289
+ "foot_ground_error": {
290
+ "max": 0.14234204590320587,
291
+ "mean": 0.07587297533660434,
292
+ "min": 0.017589636147022247,
293
+ "std": 0.021546001091225772
294
+ },
295
+ "num_samples": 86
296
+ },
297
+ "text": {
298
+ "generated": {
299
+ "matching_distance": {
300
+ "max": 1.3765006065368652,
301
+ "mean": 1.192287171995917,
302
+ "min": 1.0051730871200562,
303
+ "std": 0.07750270232156863
304
+ },
305
+ "matching_score": 1.1922872066497803,
306
+ "mean_text_rank": 8.015625,
307
+ "median_text_rank": 5.0,
308
+ "r_precision": [
309
+ 0.359375,
310
+ 0.359375,
311
+ 0.46875
312
+ ],
313
+ "retrieval_batch_size": 32,
314
+ "retrieval_num_batches": 2,
315
+ "retrieval_num_samples": 64,
316
+ "retrieval_order_policy": "manifest_order",
317
+ "retrieval_unique_datasets_per_batch": {
318
+ "max": 1.0,
319
+ "mean": 1.0,
320
+ "min": 1.0,
321
+ "std": 0.0
322
+ }
323
+ },
324
+ "reference": {
325
+ "matching_distance": {
326
+ "max": 1.3280129432678223,
327
+ "mean": 1.1493867725826974,
328
+ "min": 0.9877991080284119,
329
+ "std": 0.08739226669054043
330
+ },
331
+ "matching_score": 1.1493868827819824,
332
+ "mean_text_rank": 9.71875,
333
+ "median_text_rank": 10.0,
334
+ "r_precision": [
335
+ 0.21875,
336
+ 0.21875,
337
+ 0.234375
338
+ ],
339
+ "retrieval_batch_size": 32,
340
+ "retrieval_num_batches": 2,
341
+ "retrieval_num_samples": 64,
342
+ "retrieval_order_policy": "manifest_order",
343
+ "retrieval_unique_datasets_per_batch": {
344
+ "max": 1.0,
345
+ "mean": 1.0,
346
+ "min": 1.0,
347
+ "std": 0.0
348
+ }
349
+ }
350
+ }
351
+ },
352
+ "humanml_original_filtered_test": {
353
+ "embedding": {
354
+ "diversity_generated": 1.2619524338100758,
355
+ "motion_fid": 0.37949637116434376,
356
+ "motion_kid": {
357
+ "mean": 0.0001613574830927611,
358
+ "std": 0.0
359
+ },
360
+ "num_samples": 85
361
+ },
362
+ "num_samples": 85,
363
+ "physical": {
364
+ "body_jerk_mean": {
365
+ "max": 335.8555908203125,
366
+ "mean": 68.35952532151165,
367
+ "min": 16.2313232421875,
368
+ "std": 57.10007770878522
369
+ },
370
+ "contact_sliding_speed": {
371
+ "max": 1.2696173191070557,
372
+ "mean": 0.24481515460099806,
373
+ "min": 0.0,
374
+ "std": 0.26533379757226744
375
+ },
376
+ "foot_ground_error": {
377
+ "max": 0.22183531522750854,
378
+ "mean": 0.07665189722114626,
379
+ "min": 0.005901610944420099,
380
+ "std": 0.03276035910571276
381
+ },
382
+ "num_samples": 85
383
+ },
384
+ "text": {
385
+ "generated": {
386
+ "matching_distance": {
387
+ "max": 1.4400562047958374,
388
+ "mean": 1.119324480084812,
389
+ "min": 0.9264796376228333,
390
+ "std": 0.10692406599540519
391
+ },
392
+ "matching_score": 1.1193243265151978,
393
+ "mean_text_rank": 3.625,
394
+ "median_text_rank": 2.0,
395
+ "r_precision": [
396
+ 0.4375,
397
+ 0.59375,
398
+ 0.65625
399
+ ],
400
+ "retrieval_batch_size": 32,
401
+ "retrieval_num_batches": 2,
402
+ "retrieval_num_samples": 64,
403
+ "retrieval_order_policy": "manifest_order",
404
+ "retrieval_unique_datasets_per_batch": {
405
+ "max": 1.0,
406
+ "mean": 1.0,
407
+ "min": 1.0,
408
+ "std": 0.0
409
+ }
410
+ },
411
+ "reference": {
412
+ "matching_distance": {
413
+ "max": 1.3195350170135498,
414
+ "mean": 1.0584916128831752,
415
+ "min": 0.8765217065811157,
416
+ "std": 0.10127790732110797
417
+ },
418
+ "matching_score": 1.058491587638855,
419
+ "mean_text_rank": 2.71875,
420
+ "median_text_rank": 2.0,
421
+ "r_precision": [
422
+ 0.484375,
423
+ 0.671875,
424
+ 0.75
425
+ ],
426
+ "retrieval_batch_size": 32,
427
+ "retrieval_num_batches": 2,
428
+ "retrieval_num_samples": 64,
429
+ "retrieval_order_policy": "manifest_order",
430
+ "retrieval_unique_datasets_per_batch": {
431
+ "max": 1.0,
432
+ "mean": 1.0,
433
+ "min": 1.0,
434
+ "std": 0.0
435
+ }
436
+ }
437
+ }
438
+ },
439
+ "idea400_original_filtered_test": {
440
+ "embedding": {
441
+ "diversity_generated": 1.1603107092959224,
442
+ "motion_fid": 0.3089720542673662,
443
+ "motion_kid": {
444
+ "mean": 0.00024784405716316016,
445
+ "std": 0.0
446
+ },
447
+ "num_samples": 85
448
+ },
449
+ "num_samples": 85,
450
+ "physical": {
451
+ "body_jerk_mean": {
452
+ "max": 109.74834442138672,
453
+ "mean": 46.92937700608197,
454
+ "min": 12.396353721618652,
455
+ "std": 20.873435730428675
456
+ },
457
+ "contact_sliding_speed": {
458
+ "max": 0.8785831928253174,
459
+ "mean": 0.14769231055808418,
460
+ "min": 0.0,
461
+ "std": 0.24178951340825944
462
+ },
463
+ "foot_ground_error": {
464
+ "max": 0.2917259931564331,
465
+ "mean": 0.0424500749556019,
466
+ "min": 0.007580623496323824,
467
+ "std": 0.03828018100277618
468
+ },
469
+ "num_samples": 85
470
+ },
471
+ "text": {
472
+ "generated": {
473
+ "matching_distance": {
474
+ "max": 1.4350745677947998,
475
+ "mean": 1.1293431899126838,
476
+ "min": 0.9147061705589294,
477
+ "std": 0.10790011773020315
478
+ },
479
+ "matching_score": 1.1293431520462036,
480
+ "mean_text_rank": 7.203125,
481
+ "median_text_rank": 5.0,
482
+ "r_precision": [
483
+ 0.1875,
484
+ 0.34375,
485
+ 0.4375
486
+ ],
487
+ "retrieval_batch_size": 32,
488
+ "retrieval_num_batches": 2,
489
+ "retrieval_num_samples": 64,
490
+ "retrieval_order_policy": "manifest_order",
491
+ "retrieval_unique_datasets_per_batch": {
492
+ "max": 1.0,
493
+ "mean": 1.0,
494
+ "min": 1.0,
495
+ "std": 0.0
496
+ }
497
+ },
498
+ "reference": {
499
+ "matching_distance": {
500
+ "max": 1.3933959007263184,
501
+ "mean": 1.0893783358966602,
502
+ "min": 0.8957217931747437,
503
+ "std": 0.09054021801661577
504
+ },
505
+ "matching_score": 1.0893783569335938,
506
+ "mean_text_rank": 6.359375,
507
+ "median_text_rank": 4.0,
508
+ "r_precision": [
509
+ 0.21875,
510
+ 0.359375,
511
+ 0.46875
512
+ ],
513
+ "retrieval_batch_size": 32,
514
+ "retrieval_num_batches": 2,
515
+ "retrieval_num_samples": 64,
516
+ "retrieval_order_policy": "manifest_order",
517
+ "retrieval_unique_datasets_per_batch": {
518
+ "max": 1.0,
519
+ "mean": 1.0,
520
+ "min": 1.0,
521
+ "std": 0.0
522
+ }
523
+ }
524
+ }
525
+ },
526
+ "lafan1_original_filtered_test": {
527
+ "embedding": {
528
+ "diversity_generated": 1.0652331521765446,
529
+ "motion_fid": 0.4122437081306951,
530
+ "motion_kid": {
531
+ "mean": 0.00044010696667928784,
532
+ "std": 0.0
533
+ },
534
+ "num_samples": 85
535
+ },
536
+ "num_samples": 85,
537
+ "physical": {
538
+ "body_jerk_mean": {
539
+ "max": 314.1806945800781,
540
+ "mean": 139.9726458381204,
541
+ "min": 31.704696655273438,
542
+ "std": 71.70045395223674
543
+ },
544
+ "contact_sliding_speed": {
545
+ "max": 2.5723462104797363,
546
+ "mean": 1.074749160612769,
547
+ "min": 0.008806232362985611,
548
+ "std": 0.6605997776393984
549
+ },
550
+ "foot_ground_error": {
551
+ "max": 0.11140980571508408,
552
+ "mean": 0.021036932756781906,
553
+ "min": 0.00042374650365673006,
554
+ "std": 0.020322981130113965
555
+ },
556
+ "num_samples": 85
557
+ },
558
+ "text": {
559
+ "generated": {
560
+ "matching_distance": {
561
+ "max": 1.44863760471344,
562
+ "mean": 1.1349114824743831,
563
+ "min": 0.935748815536499,
564
+ "std": 0.10606785314295356
565
+ },
566
+ "matching_score": 1.1349115371704102,
567
+ "mean_text_rank": 8.84375,
568
+ "median_text_rank": 7.5,
569
+ "r_precision": [
570
+ 0.328125,
571
+ 0.359375,
572
+ 0.421875
573
+ ],
574
+ "retrieval_batch_size": 32,
575
+ "retrieval_num_batches": 2,
576
+ "retrieval_num_samples": 64,
577
+ "retrieval_order_policy": "manifest_order",
578
+ "retrieval_unique_datasets_per_batch": {
579
+ "max": 1.0,
580
+ "mean": 1.0,
581
+ "min": 1.0,
582
+ "std": 0.0
583
+ }
584
+ },
585
+ "reference": {
586
+ "matching_distance": {
587
+ "max": 1.4183552265167236,
588
+ "mean": 1.1475334244615891,
589
+ "min": 0.9194038510322571,
590
+ "std": 0.09896612738362433
591
+ },
592
+ "matching_score": 1.1475334167480469,
593
+ "mean_text_rank": 8.109375,
594
+ "median_text_rank": 5.0,
595
+ "r_precision": [
596
+ 0.359375,
597
+ 0.359375,
598
+ 0.375
599
+ ],
600
+ "retrieval_batch_size": 32,
601
+ "retrieval_num_batches": 2,
602
+ "retrieval_num_samples": 64,
603
+ "retrieval_order_policy": "manifest_order",
604
+ "retrieval_unique_datasets_per_batch": {
605
+ "max": 1.0,
606
+ "mean": 1.0,
607
+ "min": 1.0,
608
+ "std": 0.0
609
+ }
610
+ }
611
+ }
612
+ },
613
+ "motiongv_original_filtered_test": {
614
+ "embedding": {
615
+ "diversity_generated": 1.3252344922258321,
616
+ "motion_fid": 0.3607463810043918,
617
+ "motion_kid": {
618
+ "mean": 1.8889422511048082e-05,
619
+ "std": 0.0
620
+ },
621
+ "num_samples": 85
622
+ },
623
+ "num_samples": 85,
624
+ "physical": {
625
+ "body_jerk_mean": {
626
+ "max": 394.6324157714844,
627
+ "mean": 63.257873658572926,
628
+ "min": 10.021343231201172,
629
+ "std": 58.67747854537708
630
+ },
631
+ "contact_sliding_speed": {
632
+ "max": 0.7465254664421082,
633
+ "mean": 0.12326823623905725,
634
+ "min": 0.003070311388000846,
635
+ "std": 0.17676128678585645
636
+ },
637
+ "foot_ground_error": {
638
+ "max": 0.1117943823337555,
639
+ "mean": 0.07272094489239594,
640
+ "min": 0.012083700858056545,
641
+ "std": 0.01693819111508039
642
+ },
643
+ "num_samples": 85
644
+ },
645
+ "text": {
646
+ "generated": {
647
+ "matching_distance": {
648
+ "max": 1.4481498003005981,
649
+ "mean": 1.1457594983717974,
650
+ "min": 0.8835017085075378,
651
+ "std": 0.10666800116978434
652
+ },
653
+ "matching_score": 1.1457593441009521,
654
+ "mean_text_rank": 3.875,
655
+ "median_text_rank": 2.0,
656
+ "r_precision": [
657
+ 0.46875,
658
+ 0.59375,
659
+ 0.703125
660
+ ],
661
+ "retrieval_batch_size": 32,
662
+ "retrieval_num_batches": 2,
663
+ "retrieval_num_samples": 64,
664
+ "retrieval_order_policy": "manifest_order",
665
+ "retrieval_unique_datasets_per_batch": {
666
+ "max": 1.0,
667
+ "mean": 1.0,
668
+ "min": 1.0,
669
+ "std": 0.0
670
+ }
671
+ },
672
+ "reference": {
673
+ "matching_distance": {
674
+ "max": 1.3658052682876587,
675
+ "mean": 1.1116458219640395,
676
+ "min": 0.8830475211143494,
677
+ "std": 0.08860567172404131
678
+ },
679
+ "matching_score": 1.1116458177566528,
680
+ "mean_text_rank": 2.40625,
681
+ "median_text_rank": 1.0,
682
+ "r_precision": [
683
+ 0.609375,
684
+ 0.703125,
685
+ 0.828125
686
+ ],
687
+ "retrieval_batch_size": 32,
688
+ "retrieval_num_batches": 2,
689
+ "retrieval_num_samples": 64,
690
+ "retrieval_order_policy": "manifest_order",
691
+ "retrieval_unique_datasets_per_batch": {
692
+ "max": 1.0,
693
+ "mean": 1.0,
694
+ "min": 1.0,
695
+ "std": 0.0
696
+ }
697
+ }
698
+ }
699
+ },
700
+ "motionllama_original_filtered_test": {
701
+ "embedding": {
702
+ "diversity_generated": 1.243081346203865,
703
+ "motion_fid": 0.41212859468716423,
704
+ "motion_kid": {
705
+ "mean": 0.00039907121682780655,
706
+ "std": 0.0
707
+ },
708
+ "num_samples": 85
709
+ },
710
+ "num_samples": 85,
711
+ "physical": {
712
+ "body_jerk_mean": {
713
+ "max": 149.49559020996094,
714
+ "mean": 51.222771689471074,
715
+ "min": 11.60899543762207,
716
+ "std": 21.585048177328567
717
+ },
718
+ "contact_sliding_speed": {
719
+ "max": 0.3608182966709137,
720
+ "mean": 0.07486999163623242,
721
+ "min": 0.004085029475390911,
722
+ "std": 0.07851552799603706
723
+ },
724
+ "foot_ground_error": {
725
+ "max": 0.10084258019924164,
726
+ "mean": 0.06930571288308676,
727
+ "min": 0.012543153017759323,
728
+ "std": 0.017269924728060666
729
+ },
730
+ "num_samples": 85
731
+ },
732
+ "text": {
733
+ "generated": {
734
+ "matching_distance": {
735
+ "max": 1.3583039045333862,
736
+ "mean": 1.129518364457523,
737
+ "min": 0.907444417476654,
738
+ "std": 0.11405018318730081
739
+ },
740
+ "matching_score": 1.1295183897018433,
741
+ "mean_text_rank": 9.53125,
742
+ "median_text_rank": 7.0,
743
+ "r_precision": [
744
+ 0.171875,
745
+ 0.25,
746
+ 0.28125
747
+ ],
748
+ "retrieval_batch_size": 32,
749
+ "retrieval_num_batches": 2,
750
+ "retrieval_num_samples": 64,
751
+ "retrieval_order_policy": "manifest_order",
752
+ "retrieval_unique_datasets_per_batch": {
753
+ "max": 1.0,
754
+ "mean": 1.0,
755
+ "min": 1.0,
756
+ "std": 0.0
757
+ }
758
+ },
759
+ "reference": {
760
+ "matching_distance": {
761
+ "max": 1.2308001518249512,
762
+ "mean": 1.0292582687209635,
763
+ "min": 0.8811001777648926,
764
+ "std": 0.0877342679085498
765
+ },
766
+ "matching_score": 1.029258131980896,
767
+ "mean_text_rank": 6.921875,
768
+ "median_text_rank": 4.0,
769
+ "r_precision": [
770
+ 0.203125,
771
+ 0.3125,
772
+ 0.46875
773
+ ],
774
+ "retrieval_batch_size": 32,
775
+ "retrieval_num_batches": 2,
776
+ "retrieval_num_samples": 64,
777
+ "retrieval_order_policy": "manifest_order",
778
+ "retrieval_unique_datasets_per_batch": {
779
+ "max": 1.0,
780
+ "mean": 1.0,
781
+ "min": 1.0,
782
+ "std": 0.0
783
+ }
784
+ }
785
+ }
786
+ },
787
+ "omomo_original_filtered_test": {
788
+ "embedding": {
789
+ "diversity_generated": 1.225772022315077,
790
+ "motion_fid": 0.4057685475174302,
791
+ "motion_kid": {
792
+ "mean": 0.00022337189809062963,
793
+ "std": 0.0
794
+ },
795
+ "num_samples": 85
796
+ },
797
+ "num_samples": 85,
798
+ "physical": {
799
+ "body_jerk_mean": {
800
+ "max": 170.46762084960938,
801
+ "mean": 58.10617489534266,
802
+ "min": 16.497365951538086,
803
+ "std": 24.335758018848132
804
+ },
805
+ "contact_sliding_speed": {
806
+ "max": 1.117565631866455,
807
+ "mean": 0.3834191459504997,
808
+ "min": 0.005078582558780909,
809
+ "std": 0.3170930006320161
810
+ },
811
+ "foot_ground_error": {
812
+ "max": 0.1168566644191742,
813
+ "mean": 0.016199057431453293,
814
+ "min": 0.002339268336072564,
815
+ "std": 0.013331415234509801
816
+ },
817
+ "num_samples": 85
818
+ },
819
+ "text": {
820
+ "generated": {
821
+ "matching_distance": {
822
+ "max": 1.4282751083374023,
823
+ "mean": 1.1995593414587133,
824
+ "min": 0.970937192440033,
825
+ "std": 0.09437652461416668
826
+ },
827
+ "matching_score": 1.1995593309402466,
828
+ "mean_text_rank": 9.390625,
829
+ "median_text_rank": 7.0,
830
+ "r_precision": [
831
+ 0.234375,
832
+ 0.328125,
833
+ 0.34375
834
+ ],
835
+ "retrieval_batch_size": 32,
836
+ "retrieval_num_batches": 2,
837
+ "retrieval_num_samples": 64,
838
+ "retrieval_order_policy": "manifest_order",
839
+ "retrieval_unique_datasets_per_batch": {
840
+ "max": 1.0,
841
+ "mean": 1.0,
842
+ "min": 1.0,
843
+ "std": 0.0
844
+ }
845
+ },
846
+ "reference": {
847
+ "matching_distance": {
848
+ "max": 1.3254883289337158,
849
+ "mean": 1.1843213705455555,
850
+ "min": 0.958416759967804,
851
+ "std": 0.07360095182352315
852
+ },
853
+ "matching_score": 1.1843212842941284,
854
+ "mean_text_rank": 7.484375,
855
+ "median_text_rank": 5.5,
856
+ "r_precision": [
857
+ 0.140625,
858
+ 0.328125,
859
+ 0.40625
860
+ ],
861
+ "retrieval_batch_size": 32,
862
+ "retrieval_num_batches": 2,
863
+ "retrieval_num_samples": 64,
864
+ "retrieval_order_policy": "manifest_order",
865
+ "retrieval_unique_datasets_per_batch": {
866
+ "max": 1.0,
867
+ "mean": 1.0,
868
+ "min": 1.0,
869
+ "std": 0.0
870
+ }
871
+ }
872
+ }
873
+ },
874
+ "permo_original_filtered_test": {
875
+ "embedding": {
876
+ "diversity_generated": 1.2086325566904998,
877
+ "motion_fid": 0.4148669497066708,
878
+ "motion_kid": {
879
+ "mean": 0.0003991512469281311,
880
+ "std": 0.0
881
+ },
882
+ "num_samples": 85
883
+ },
884
+ "num_samples": 85,
885
+ "physical": {
886
+ "body_jerk_mean": {
887
+ "max": 240.15951538085938,
888
+ "mean": 89.35240687202005,
889
+ "min": 24.114988327026367,
890
+ "std": 42.25908942517282
891
+ },
892
+ "contact_sliding_speed": {
893
+ "max": 1.3213852643966675,
894
+ "mean": 0.38401777897687517,
895
+ "min": 0.0,
896
+ "std": 0.3190618849745317
897
+ },
898
+ "foot_ground_error": {
899
+ "max": 0.2258841097354889,
900
+ "mean": 0.036138936230803234,
901
+ "min": 0.008152862079441547,
902
+ "std": 0.02827083711142457
903
+ },
904
+ "num_samples": 85
905
+ },
906
+ "text": {
907
+ "generated": {
908
+ "matching_distance": {
909
+ "max": 1.412964940071106,
910
+ "mean": 1.104096428085776,
911
+ "min": 0.8289653062820435,
912
+ "std": 0.12450774868152055
913
+ },
914
+ "matching_score": 1.1040964126586914,
915
+ "mean_text_rank": 5.796875,
916
+ "median_text_rank": 4.0,
917
+ "r_precision": [
918
+ 0.15625,
919
+ 0.359375,
920
+ 0.4375
921
+ ],
922
+ "retrieval_batch_size": 32,
923
+ "retrieval_num_batches": 2,
924
+ "retrieval_num_samples": 64,
925
+ "retrieval_order_policy": "manifest_order",
926
+ "retrieval_unique_datasets_per_batch": {
927
+ "max": 1.0,
928
+ "mean": 1.0,
929
+ "min": 1.0,
930
+ "std": 0.0
931
+ }
932
+ },
933
+ "reference": {
934
+ "matching_distance": {
935
+ "max": 1.2057435512542725,
936
+ "mean": 1.0011776215889874,
937
+ "min": 0.8405618071556091,
938
+ "std": 0.0876302018200968
939
+ },
940
+ "matching_score": 1.0011776685714722,
941
+ "mean_text_rank": 4.84375,
942
+ "median_text_rank": 4.0,
943
+ "r_precision": [
944
+ 0.171875,
945
+ 0.328125,
946
+ 0.46875
947
+ ],
948
+ "retrieval_batch_size": 32,
949
+ "retrieval_num_batches": 2,
950
+ "retrieval_num_samples": 64,
951
+ "retrieval_order_policy": "manifest_order",
952
+ "retrieval_unique_datasets_per_batch": {
953
+ "max": 1.0,
954
+ "mean": 1.0,
955
+ "min": 1.0,
956
+ "std": 0.0
957
+ }
958
+ }
959
+ }
960
+ },
961
+ "snapmogen_original_filtered_test": {
962
+ "embedding": {
963
+ "diversity_generated": 1.1717517212600463,
964
+ "motion_fid": 0.3023945940485344,
965
+ "motion_kid": {
966
+ "mean": 0.0002671196633574624,
967
+ "std": 0.0
968
+ },
969
+ "num_samples": 85
970
+ },
971
+ "num_samples": 85,
972
+ "physical": {
973
+ "body_jerk_mean": {
974
+ "max": 183.62680053710938,
975
+ "mean": 57.84897658404182,
976
+ "min": 16.209308624267578,
977
+ "std": 38.09984754580235
978
+ },
979
+ "contact_sliding_speed": {
980
+ "max": 1.0688523054122925,
981
+ "mean": 0.1868419358452015,
982
+ "min": 0.0041826097294688225,
983
+ "std": 0.20431294637395492
984
+ },
985
+ "foot_ground_error": {
986
+ "max": 0.14229609072208405,
987
+ "mean": 0.034169711305431144,
988
+ "min": 0.0016218408709391952,
989
+ "std": 0.02866694918161823
990
+ },
991
+ "num_samples": 85
992
+ },
993
+ "text": {
994
+ "generated": {
995
+ "matching_distance": {
996
+ "max": 1.3117051124572754,
997
+ "mean": 1.0969315627041984,
998
+ "min": 0.9296107888221741,
999
+ "std": 0.08918784110822944
1000
+ },
1001
+ "matching_score": 1.0969316959381104,
1002
+ "mean_text_rank": 4.265625,
1003
+ "median_text_rank": 3.0,
1004
+ "r_precision": [
1005
+ 0.359375,
1006
+ 0.46875,
1007
+ 0.65625
1008
+ ],
1009
+ "retrieval_batch_size": 32,
1010
+ "retrieval_num_batches": 2,
1011
+ "retrieval_num_samples": 64,
1012
+ "retrieval_order_policy": "manifest_order",
1013
+ "retrieval_unique_datasets_per_batch": {
1014
+ "max": 1.0,
1015
+ "mean": 1.0,
1016
+ "min": 1.0,
1017
+ "std": 0.0
1018
+ }
1019
+ },
1020
+ "reference": {
1021
+ "matching_distance": {
1022
+ "max": 1.2125790119171143,
1023
+ "mean": 1.0510956680073458,
1024
+ "min": 0.8823143839836121,
1025
+ "std": 0.08172085751280951
1026
+ },
1027
+ "matching_score": 1.051095724105835,
1028
+ "mean_text_rank": 3.140625,
1029
+ "median_text_rank": 2.0,
1030
+ "r_precision": [
1031
+ 0.46875,
1032
+ 0.671875,
1033
+ 0.765625
1034
+ ],
1035
+ "retrieval_batch_size": 32,
1036
+ "retrieval_num_batches": 2,
1037
+ "retrieval_num_samples": 64,
1038
+ "retrieval_order_policy": "manifest_order",
1039
+ "retrieval_unique_datasets_per_batch": {
1040
+ "max": 1.0,
1041
+ "mean": 1.0,
1042
+ "min": 1.0,
1043
+ "std": 0.0
1044
+ }
1045
+ }
1046
+ }
1047
+ }
1048
+ },
1049
+ "embedding_frames": 120,
1050
+ "embedding_metrics": {
1051
+ "diversity_generated": 1.3800554760633976,
1052
+ "embedding_frames": 120,
1053
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
1054
+ "matching_score_generated": 1.1404733657836914,
1055
+ "matching_score_reference": 1.0944490432739258,
1056
+ "motion_fid": 0.10967632126673776,
1057
+ "motion_key": "body_pos_local",
1058
+ "motion_kid": {
1059
+ "mean": 7.427152145522697e-05,
1060
+ "std": 0.0
1061
+ },
1062
+ "multimodality": null,
1063
+ "num_samples": 1024,
1064
+ "r_precision_generated": [
1065
+ 0.5869140625,
1066
+ 0.7333984375,
1067
+ 0.818359375
1068
+ ],
1069
+ "r_precision_reference": [
1070
+ 0.6611328125,
1071
+ 0.82421875,
1072
+ 0.8916015625
1073
+ ],
1074
+ "text_cache_dir": "${OMG_REPO_ROOT}/outputs/evaluator_text_cache/filtered_original_g1_tmr_bodypos_60f",
1075
+ "text_encoder_model": "${OMG_REPO_ROOT}/models/t5-3b-local",
1076
+ "text_generated": {
1077
+ "matching_distance": {
1078
+ "max": 1.5543382167816162,
1079
+ "mean": 1.1404733364470303,
1080
+ "min": 0.8289653062820435,
1081
+ "std": 0.10747345553020096
1082
+ },
1083
+ "matching_score": 1.1404733657836914,
1084
+ "mean_text_rank": 2.576171875,
1085
+ "median_text_rank": 1.0,
1086
+ "r_precision": [
1087
+ 0.5869140625,
1088
+ 0.7333984375,
1089
+ 0.818359375
1090
+ ],
1091
+ "retrieval_batch_size": 32,
1092
+ "retrieval_num_batches": 32,
1093
+ "retrieval_num_samples": 1024,
1094
+ "retrieval_order_policy": "stratified_by_dataset",
1095
+ "retrieval_unique_datasets_per_batch": {
1096
+ "max": 12.0,
1097
+ "mean": 11.9375,
1098
+ "min": 10.0,
1099
+ "std": 0.34798527267687634
1100
+ }
1101
+ },
1102
+ "text_reference": {
1103
+ "matching_distance": {
1104
+ "max": 1.4848475456237793,
1105
+ "mean": 1.0944489734247327,
1106
+ "min": 0.8405618071556091,
1107
+ "std": 0.10523510496847523
1108
+ },
1109
+ "matching_score": 1.0944490432739258,
1110
+ "mean_text_rank": 1.888671875,
1111
+ "median_text_rank": 1.0,
1112
+ "r_precision": [
1113
+ 0.6611328125,
1114
+ 0.82421875,
1115
+ 0.8916015625
1116
+ ],
1117
+ "retrieval_batch_size": 32,
1118
+ "retrieval_num_batches": 32,
1119
+ "retrieval_num_samples": 1024,
1120
+ "retrieval_order_policy": "stratified_by_dataset",
1121
+ "retrieval_unique_datasets_per_batch": {
1122
+ "max": 12.0,
1123
+ "mean": 11.9375,
1124
+ "min": 10.0,
1125
+ "std": 0.34798527267687634
1126
+ }
1127
+ },
1128
+ "text_retrieval_batch_size": 32
1129
+ },
1130
+ "exp": "300m",
1131
+ "generated_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/generated_qpos_raw.npz",
1132
+ "label": "OmniMotion 300M mixed-modal scaling s050000",
1133
+ "num_frames": 120,
1134
+ "num_samples": 1024,
1135
+ "physical_metrics": {
1136
+ "body_jerk_mean": {
1137
+ "max": 394.6324157714844,
1138
+ "mean": 74.01262681838125,
1139
+ "min": 10.021343231201172,
1140
+ "std": 53.226579084617505
1141
+ },
1142
+ "contact_height_threshold": 0.12,
1143
+ "contact_penetration_tolerance": 0.02,
1144
+ "contact_sliding_speed": {
1145
+ "max": 2.5723462104797363,
1146
+ "mean": 0.35662794705444867,
1147
+ "min": 0.0,
1148
+ "std": 0.4447682159597098
1149
+ },
1150
+ "foot_ground_error": {
1151
+ "max": 0.2917259931564331,
1152
+ "mean": 0.04151647091404698,
1153
+ "min": 0.00042374650365673006,
1154
+ "std": 0.03340117411035358
1155
+ },
1156
+ "frames": 120,
1157
+ "num_samples": 1024
1158
+ },
1159
+ "real_motion_metrics": {
1160
+ "diversity_real": 1.3768246488138,
1161
+ "matching_distance_real": {
1162
+ "max": 1.4848475456237793,
1163
+ "mean": 1.0944489734247327,
1164
+ "min": 0.8405618071556091,
1165
+ "std": 0.10523510496847523
1166
+ },
1167
+ "matching_score_real": 1.0944490432739258,
1168
+ "motion_fid_reference_eval": 0.0,
1169
+ "motion_fid_split_half": 0.0,
1170
+ "motion_kid_reference_eval": {
1171
+ "mean": -1.0749719189018236e-05,
1172
+ "std": 0.0
1173
+ },
1174
+ "motion_kid_split_half": {
1175
+ "mean": -1.0749719189018236e-05,
1176
+ "std": 0.0
1177
+ },
1178
+ "multimodality": "-",
1179
+ "num_reference_samples": 1024,
1180
+ "num_samples": 1024,
1181
+ "r_precision_real": [
1182
+ 0.3740234375,
1183
+ 0.5009765625,
1184
+ 0.5810546875
1185
+ ],
1186
+ "text_real": {
1187
+ "matching_distance": {
1188
+ "max": 1.4848475456237793,
1189
+ "mean": 1.0944489734247327,
1190
+ "min": 0.8405618071556091,
1191
+ "std": 0.10523510496847523
1192
+ },
1193
+ "matching_score": 1.0944490432739258,
1194
+ "mean_text_rank": 5.232421875,
1195
+ "median_text_rank": 2.0,
1196
+ "r_precision": [
1197
+ 0.3740234375,
1198
+ 0.5009765625,
1199
+ 0.5810546875
1200
+ ],
1201
+ "retrieval_batch_size": 32,
1202
+ "retrieval_num_batches": 32,
1203
+ "retrieval_num_samples": 1024,
1204
+ "retrieval_order_policy": "manifest_order",
1205
+ "retrieval_unique_datasets_per_batch": null
1206
+ }
1207
+ },
1208
+ "reference_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/reference_qpos.npz",
1209
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/sample_metrics.jsonl",
1210
+ "sample_rankings_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/sample_rankings.json",
1211
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_samples/mixed_modalities_all_v1/text_test_1024.jsonl",
1212
+ "split": "test",
1213
+ "tracker_executed": {
1214
+ "action_clip": 10.0,
1215
+ "artifact_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/tracker_executed/tracker_executed_qpos.npz",
1216
+ "benchmark_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/tracker_executed/benchmark_tracker.json",
1217
+ "control_substeps": 10,
1218
+ "enabled": true,
1219
+ "holomotion_onnx": "${OMG_MODELS_ROOT}/HoloMotion/HoloMotion_v1.2/holomotion_v1.2_motion_tracking_model/exported/model.onnx",
1220
+ "metrics": {
1221
+ "e_acc": {
1222
+ "max": 5.753578186035156,
1223
+ "mean": 1.3608852332108654,
1224
+ "min": 0.19200170040130615,
1225
+ "std": 1.0193065179674143
1226
+ },
1227
+ "e_vel": {
1228
+ "max": 27.857595443725586,
1229
+ "mean": 4.456644410674926,
1230
+ "min": 0.6209603548049927,
1231
+ "std": 3.3743479287514506
1232
+ },
1233
+ "enabled": true,
1234
+ "g_mpjpe": {
1235
+ "max": 1602.4024658203125,
1236
+ "mean": 130.28732685372233,
1237
+ "min": 18.030338287353516,
1238
+ "std": 129.4890607920233
1239
+ },
1240
+ "mpjpe": {
1241
+ "max": 316.797119140625,
1242
+ "mean": 45.63376460596919,
1243
+ "min": 13.653532981872559,
1244
+ "std": 27.06205114115315
1245
+ },
1246
+ "num_samples": 1024,
1247
+ "root_index": 0,
1248
+ "target_fps": 50.0
1249
+ },
1250
+ "metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/tracker_executed/tracker_metrics.json",
1251
+ "num_generated_samples": 1024,
1252
+ "num_samples": 1024,
1253
+ "providers": "TensorrtExecutionProvider,CUDAExecutionProvider",
1254
+ "robot_xml": null,
1255
+ "root_index": 0,
1256
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/tracker_executed/tracker_sample_metrics.jsonl",
1257
+ "selected_indices": [
1258
+ 0,
1259
+ 1,
1260
+ 2,
1261
+ 3,
1262
+ 4,
1263
+ 5,
1264
+ 6,
1265
+ 7,
1266
+ 8,
1267
+ 9,
1268
+ 10,
1269
+ 11,
1270
+ 12,
1271
+ 13,
1272
+ 14,
1273
+ 15,
1274
+ 16,
1275
+ 17,
1276
+ 18,
1277
+ 19,
1278
+ 20,
1279
+ 21,
1280
+ 22,
1281
+ 23,
1282
+ 24,
1283
+ 25,
1284
+ 26,
1285
+ 27,
1286
+ 28,
1287
+ 29,
1288
+ 30,
1289
+ 31,
1290
+ 32,
1291
+ 33,
1292
+ 34,
1293
+ 35,
1294
+ 36,
1295
+ 37,
1296
+ 38,
1297
+ 39,
1298
+ 40,
1299
+ 41,
1300
+ 42,
1301
+ 43,
1302
+ 44,
1303
+ 45,
1304
+ 46,
1305
+ 47,
1306
+ 48,
1307
+ 49,
1308
+ 50,
1309
+ 51,
1310
+ 52,
1311
+ 53,
1312
+ 54,
1313
+ 55,
1314
+ 56,
1315
+ 57,
1316
+ 58,
1317
+ 59,
1318
+ 60,
1319
+ 61,
1320
+ 62,
1321
+ 63,
1322
+ 64,
1323
+ 65,
1324
+ 66,
1325
+ 67,
1326
+ 68,
1327
+ 69,
1328
+ 70,
1329
+ 71,
1330
+ 72,
1331
+ 73,
1332
+ 74,
1333
+ 75,
1334
+ 76,
1335
+ 77,
1336
+ 78,
1337
+ 79,
1338
+ 80,
1339
+ 81,
1340
+ 82,
1341
+ 83,
1342
+ 84,
1343
+ 85,
1344
+ 86,
1345
+ 87,
1346
+ 88,
1347
+ 89,
1348
+ 90,
1349
+ 91,
1350
+ 92,
1351
+ 93,
1352
+ 94,
1353
+ 95,
1354
+ 96,
1355
+ 97,
1356
+ 98,
1357
+ 99,
1358
+ 100,
1359
+ 101,
1360
+ 102,
1361
+ 103,
1362
+ 104,
1363
+ 105,
1364
+ 106,
1365
+ 107,
1366
+ 108,
1367
+ 109,
1368
+ 110,
1369
+ 111,
1370
+ 112,
1371
+ 113,
1372
+ 114,
1373
+ 115,
1374
+ 116,
1375
+ 117,
1376
+ 118,
1377
+ 119,
1378
+ 120,
1379
+ 121,
1380
+ 122,
1381
+ 123,
1382
+ 124,
1383
+ 125,
1384
+ 126,
1385
+ 127,
1386
+ 128,
1387
+ 129,
1388
+ 130,
1389
+ 131,
1390
+ 132,
1391
+ 133,
1392
+ 134,
1393
+ 135,
1394
+ 136,
1395
+ 137,
1396
+ 138,
1397
+ 139,
1398
+ 140,
1399
+ 141,
1400
+ 142,
1401
+ 143,
1402
+ 144,
1403
+ 145,
1404
+ 146,
1405
+ 147,
1406
+ 148,
1407
+ 149,
1408
+ 150,
1409
+ 151,
1410
+ 152,
1411
+ 153,
1412
+ 154,
1413
+ 155,
1414
+ 156,
1415
+ 157,
1416
+ 158,
1417
+ 159,
1418
+ 160,
1419
+ 161,
1420
+ 162,
1421
+ 163,
1422
+ 164,
1423
+ 165,
1424
+ 166,
1425
+ 167,
1426
+ 168,
1427
+ 169,
1428
+ 170,
1429
+ 171,
1430
+ 172,
1431
+ 173,
1432
+ 174,
1433
+ 175,
1434
+ 176,
1435
+ 177,
1436
+ 178,
1437
+ 179,
1438
+ 180,
1439
+ 181,
1440
+ 182,
1441
+ 183,
1442
+ 184,
1443
+ 185,
1444
+ 186,
1445
+ 187,
1446
+ 188,
1447
+ 189,
1448
+ 190,
1449
+ 191,
1450
+ 192,
1451
+ 193,
1452
+ 194,
1453
+ 195,
1454
+ 196,
1455
+ 197,
1456
+ 198,
1457
+ 199,
1458
+ 200,
1459
+ 201,
1460
+ 202,
1461
+ 203,
1462
+ 204,
1463
+ 205,
1464
+ 206,
1465
+ 207,
1466
+ 208,
1467
+ 209,
1468
+ 210,
1469
+ 211,
1470
+ 212,
1471
+ 213,
1472
+ 214,
1473
+ 215,
1474
+ 216,
1475
+ 217,
1476
+ 218,
1477
+ 219,
1478
+ 220,
1479
+ 221,
1480
+ 222,
1481
+ 223,
1482
+ 224,
1483
+ 225,
1484
+ 226,
1485
+ 227,
1486
+ 228,
1487
+ 229,
1488
+ 230,
1489
+ 231,
1490
+ 232,
1491
+ 233,
1492
+ 234,
1493
+ 235,
1494
+ 236,
1495
+ 237,
1496
+ 238,
1497
+ 239,
1498
+ 240,
1499
+ 241,
1500
+ 242,
1501
+ 243,
1502
+ 244,
1503
+ 245,
1504
+ 246,
1505
+ 247,
1506
+ 248,
1507
+ 249,
1508
+ 250,
1509
+ 251,
1510
+ 252,
1511
+ 253,
1512
+ 254,
1513
+ 255,
1514
+ 256,
1515
+ 257,
1516
+ 258,
1517
+ 259,
1518
+ 260,
1519
+ 261,
1520
+ 262,
1521
+ 263,
1522
+ 264,
1523
+ 265,
1524
+ 266,
1525
+ 267,
1526
+ 268,
1527
+ 269,
1528
+ 270,
1529
+ 271,
1530
+ 272,
1531
+ 273,
1532
+ 274,
1533
+ 275,
1534
+ 276,
1535
+ 277,
1536
+ 278,
1537
+ 279,
1538
+ 280,
1539
+ 281,
1540
+ 282,
1541
+ 283,
1542
+ 284,
1543
+ 285,
1544
+ 286,
1545
+ 287,
1546
+ 288,
1547
+ 289,
1548
+ 290,
1549
+ 291,
1550
+ 292,
1551
+ 293,
1552
+ 294,
1553
+ 295,
1554
+ 296,
1555
+ 297,
1556
+ 298,
1557
+ 299,
1558
+ 300,
1559
+ 301,
1560
+ 302,
1561
+ 303,
1562
+ 304,
1563
+ 305,
1564
+ 306,
1565
+ 307,
1566
+ 308,
1567
+ 309,
1568
+ 310,
1569
+ 311,
1570
+ 312,
1571
+ 313,
1572
+ 314,
1573
+ 315,
1574
+ 316,
1575
+ 317,
1576
+ 318,
1577
+ 319,
1578
+ 320,
1579
+ 321,
1580
+ 322,
1581
+ 323,
1582
+ 324,
1583
+ 325,
1584
+ 326,
1585
+ 327,
1586
+ 328,
1587
+ 329,
1588
+ 330,
1589
+ 331,
1590
+ 332,
1591
+ 333,
1592
+ 334,
1593
+ 335,
1594
+ 336,
1595
+ 337,
1596
+ 338,
1597
+ 339,
1598
+ 340,
1599
+ 341,
1600
+ 342,
1601
+ 343,
1602
+ 344,
1603
+ 345,
1604
+ 346,
1605
+ 347,
1606
+ 348,
1607
+ 349,
1608
+ 350,
1609
+ 351,
1610
+ 352,
1611
+ 353,
1612
+ 354,
1613
+ 355,
1614
+ 356,
1615
+ 357,
1616
+ 358,
1617
+ 359,
1618
+ 360,
1619
+ 361,
1620
+ 362,
1621
+ 363,
1622
+ 364,
1623
+ 365,
1624
+ 366,
1625
+ 367,
1626
+ 368,
1627
+ 369,
1628
+ 370,
1629
+ 371,
1630
+ 372,
1631
+ 373,
1632
+ 374,
1633
+ 375,
1634
+ 376,
1635
+ 377,
1636
+ 378,
1637
+ 379,
1638
+ 380,
1639
+ 381,
1640
+ 382,
1641
+ 383,
1642
+ 384,
1643
+ 385,
1644
+ 386,
1645
+ 387,
1646
+ 388,
1647
+ 389,
1648
+ 390,
1649
+ 391,
1650
+ 392,
1651
+ 393,
1652
+ 394,
1653
+ 395,
1654
+ 396,
1655
+ 397,
1656
+ 398,
1657
+ 399,
1658
+ 400,
1659
+ 401,
1660
+ 402,
1661
+ 403,
1662
+ 404,
1663
+ 405,
1664
+ 406,
1665
+ 407,
1666
+ 408,
1667
+ 409,
1668
+ 410,
1669
+ 411,
1670
+ 412,
1671
+ 413,
1672
+ 414,
1673
+ 415,
1674
+ 416,
1675
+ 417,
1676
+ 418,
1677
+ 419,
1678
+ 420,
1679
+ 421,
1680
+ 422,
1681
+ 423,
1682
+ 424,
1683
+ 425,
1684
+ 426,
1685
+ 427,
1686
+ 428,
1687
+ 429,
1688
+ 430,
1689
+ 431,
1690
+ 432,
1691
+ 433,
1692
+ 434,
1693
+ 435,
1694
+ 436,
1695
+ 437,
1696
+ 438,
1697
+ 439,
1698
+ 440,
1699
+ 441,
1700
+ 442,
1701
+ 443,
1702
+ 444,
1703
+ 445,
1704
+ 446,
1705
+ 447,
1706
+ 448,
1707
+ 449,
1708
+ 450,
1709
+ 451,
1710
+ 452,
1711
+ 453,
1712
+ 454,
1713
+ 455,
1714
+ 456,
1715
+ 457,
1716
+ 458,
1717
+ 459,
1718
+ 460,
1719
+ 461,
1720
+ 462,
1721
+ 463,
1722
+ 464,
1723
+ 465,
1724
+ 466,
1725
+ 467,
1726
+ 468,
1727
+ 469,
1728
+ 470,
1729
+ 471,
1730
+ 472,
1731
+ 473,
1732
+ 474,
1733
+ 475,
1734
+ 476,
1735
+ 477,
1736
+ 478,
1737
+ 479,
1738
+ 480,
1739
+ 481,
1740
+ 482,
1741
+ 483,
1742
+ 484,
1743
+ 485,
1744
+ 486,
1745
+ 487,
1746
+ 488,
1747
+ 489,
1748
+ 490,
1749
+ 491,
1750
+ 492,
1751
+ 493,
1752
+ 494,
1753
+ 495,
1754
+ 496,
1755
+ 497,
1756
+ 498,
1757
+ 499,
1758
+ 500,
1759
+ 501,
1760
+ 502,
1761
+ 503,
1762
+ 504,
1763
+ 505,
1764
+ 506,
1765
+ 507,
1766
+ 508,
1767
+ 509,
1768
+ 510,
1769
+ 511,
1770
+ 512,
1771
+ 513,
1772
+ 514,
1773
+ 515,
1774
+ 516,
1775
+ 517,
1776
+ 518,
1777
+ 519,
1778
+ 520,
1779
+ 521,
1780
+ 522,
1781
+ 523,
1782
+ 524,
1783
+ 525,
1784
+ 526,
1785
+ 527,
1786
+ 528,
1787
+ 529,
1788
+ 530,
1789
+ 531,
1790
+ 532,
1791
+ 533,
1792
+ 534,
1793
+ 535,
1794
+ 536,
1795
+ 537,
1796
+ 538,
1797
+ 539,
1798
+ 540,
1799
+ 541,
1800
+ 542,
1801
+ 543,
1802
+ 544,
1803
+ 545,
1804
+ 546,
1805
+ 547,
1806
+ 548,
1807
+ 549,
1808
+ 550,
1809
+ 551,
1810
+ 552,
1811
+ 553,
1812
+ 554,
1813
+ 555,
1814
+ 556,
1815
+ 557,
1816
+ 558,
1817
+ 559,
1818
+ 560,
1819
+ 561,
1820
+ 562,
1821
+ 563,
1822
+ 564,
1823
+ 565,
1824
+ 566,
1825
+ 567,
1826
+ 568,
1827
+ 569,
1828
+ 570,
1829
+ 571,
1830
+ 572,
1831
+ 573,
1832
+ 574,
1833
+ 575,
1834
+ 576,
1835
+ 577,
1836
+ 578,
1837
+ 579,
1838
+ 580,
1839
+ 581,
1840
+ 582,
1841
+ 583,
1842
+ 584,
1843
+ 585,
1844
+ 586,
1845
+ 587,
1846
+ 588,
1847
+ 589,
1848
+ 590,
1849
+ 591,
1850
+ 592,
1851
+ 593,
1852
+ 594,
1853
+ 595,
1854
+ 596,
1855
+ 597,
1856
+ 598,
1857
+ 599,
1858
+ 600,
1859
+ 601,
1860
+ 602,
1861
+ 603,
1862
+ 604,
1863
+ 605,
1864
+ 606,
1865
+ 607,
1866
+ 608,
1867
+ 609,
1868
+ 610,
1869
+ 611,
1870
+ 612,
1871
+ 613,
1872
+ 614,
1873
+ 615,
1874
+ 616,
1875
+ 617,
1876
+ 618,
1877
+ 619,
1878
+ 620,
1879
+ 621,
1880
+ 622,
1881
+ 623,
1882
+ 624,
1883
+ 625,
1884
+ 626,
1885
+ 627,
1886
+ 628,
1887
+ 629,
1888
+ 630,
1889
+ 631,
1890
+ 632,
1891
+ 633,
1892
+ 634,
1893
+ 635,
1894
+ 636,
1895
+ 637,
1896
+ 638,
1897
+ 639,
1898
+ 640,
1899
+ 641,
1900
+ 642,
1901
+ 643,
1902
+ 644,
1903
+ 645,
1904
+ 646,
1905
+ 647,
1906
+ 648,
1907
+ 649,
1908
+ 650,
1909
+ 651,
1910
+ 652,
1911
+ 653,
1912
+ 654,
1913
+ 655,
1914
+ 656,
1915
+ 657,
1916
+ 658,
1917
+ 659,
1918
+ 660,
1919
+ 661,
1920
+ 662,
1921
+ 663,
1922
+ 664,
1923
+ 665,
1924
+ 666,
1925
+ 667,
1926
+ 668,
1927
+ 669,
1928
+ 670,
1929
+ 671,
1930
+ 672,
1931
+ 673,
1932
+ 674,
1933
+ 675,
1934
+ 676,
1935
+ 677,
1936
+ 678,
1937
+ 679,
1938
+ 680,
1939
+ 681,
1940
+ 682,
1941
+ 683,
1942
+ 684,
1943
+ 685,
1944
+ 686,
1945
+ 687,
1946
+ 688,
1947
+ 689,
1948
+ 690,
1949
+ 691,
1950
+ 692,
1951
+ 693,
1952
+ 694,
1953
+ 695,
1954
+ 696,
1955
+ 697,
1956
+ 698,
1957
+ 699,
1958
+ 700,
1959
+ 701,
1960
+ 702,
1961
+ 703,
1962
+ 704,
1963
+ 705,
1964
+ 706,
1965
+ 707,
1966
+ 708,
1967
+ 709,
1968
+ 710,
1969
+ 711,
1970
+ 712,
1971
+ 713,
1972
+ 714,
1973
+ 715,
1974
+ 716,
1975
+ 717,
1976
+ 718,
1977
+ 719,
1978
+ 720,
1979
+ 721,
1980
+ 722,
1981
+ 723,
1982
+ 724,
1983
+ 725,
1984
+ 726,
1985
+ 727,
1986
+ 728,
1987
+ 729,
1988
+ 730,
1989
+ 731,
1990
+ 732,
1991
+ 733,
1992
+ 734,
1993
+ 735,
1994
+ 736,
1995
+ 737,
1996
+ 738,
1997
+ 739,
1998
+ 740,
1999
+ 741,
2000
+ 742,
2001
+ 743,
2002
+ 744,
2003
+ 745,
2004
+ 746,
2005
+ 747,
2006
+ 748,
2007
+ 749,
2008
+ 750,
2009
+ 751,
2010
+ 752,
2011
+ 753,
2012
+ 754,
2013
+ 755,
2014
+ 756,
2015
+ 757,
2016
+ 758,
2017
+ 759,
2018
+ 760,
2019
+ 761,
2020
+ 762,
2021
+ 763,
2022
+ 764,
2023
+ 765,
2024
+ 766,
2025
+ 767,
2026
+ 768,
2027
+ 769,
2028
+ 770,
2029
+ 771,
2030
+ 772,
2031
+ 773,
2032
+ 774,
2033
+ 775,
2034
+ 776,
2035
+ 777,
2036
+ 778,
2037
+ 779,
2038
+ 780,
2039
+ 781,
2040
+ 782,
2041
+ 783,
2042
+ 784,
2043
+ 785,
2044
+ 786,
2045
+ 787,
2046
+ 788,
2047
+ 789,
2048
+ 790,
2049
+ 791,
2050
+ 792,
2051
+ 793,
2052
+ 794,
2053
+ 795,
2054
+ 796,
2055
+ 797,
2056
+ 798,
2057
+ 799,
2058
+ 800,
2059
+ 801,
2060
+ 802,
2061
+ 803,
2062
+ 804,
2063
+ 805,
2064
+ 806,
2065
+ 807,
2066
+ 808,
2067
+ 809,
2068
+ 810,
2069
+ 811,
2070
+ 812,
2071
+ 813,
2072
+ 814,
2073
+ 815,
2074
+ 816,
2075
+ 817,
2076
+ 818,
2077
+ 819,
2078
+ 820,
2079
+ 821,
2080
+ 822,
2081
+ 823,
2082
+ 824,
2083
+ 825,
2084
+ 826,
2085
+ 827,
2086
+ 828,
2087
+ 829,
2088
+ 830,
2089
+ 831,
2090
+ 832,
2091
+ 833,
2092
+ 834,
2093
+ 835,
2094
+ 836,
2095
+ 837,
2096
+ 838,
2097
+ 839,
2098
+ 840,
2099
+ 841,
2100
+ 842,
2101
+ 843,
2102
+ 844,
2103
+ 845,
2104
+ 846,
2105
+ 847,
2106
+ 848,
2107
+ 849,
2108
+ 850,
2109
+ 851,
2110
+ 852,
2111
+ 853,
2112
+ 854,
2113
+ 855,
2114
+ 856,
2115
+ 857,
2116
+ 858,
2117
+ 859,
2118
+ 860,
2119
+ 861,
2120
+ 862,
2121
+ 863,
2122
+ 864,
2123
+ 865,
2124
+ 866,
2125
+ 867,
2126
+ 868,
2127
+ 869,
2128
+ 870,
2129
+ 871,
2130
+ 872,
2131
+ 873,
2132
+ 874,
2133
+ 875,
2134
+ 876,
2135
+ 877,
2136
+ 878,
2137
+ 879,
2138
+ 880,
2139
+ 881,
2140
+ 882,
2141
+ 883,
2142
+ 884,
2143
+ 885,
2144
+ 886,
2145
+ 887,
2146
+ 888,
2147
+ 889,
2148
+ 890,
2149
+ 891,
2150
+ 892,
2151
+ 893,
2152
+ 894,
2153
+ 895,
2154
+ 896,
2155
+ 897,
2156
+ 898,
2157
+ 899,
2158
+ 900,
2159
+ 901,
2160
+ 902,
2161
+ 903,
2162
+ 904,
2163
+ 905,
2164
+ 906,
2165
+ 907,
2166
+ 908,
2167
+ 909,
2168
+ 910,
2169
+ 911,
2170
+ 912,
2171
+ 913,
2172
+ 914,
2173
+ 915,
2174
+ 916,
2175
+ 917,
2176
+ 918,
2177
+ 919,
2178
+ 920,
2179
+ 921,
2180
+ 922,
2181
+ 923,
2182
+ 924,
2183
+ 925,
2184
+ 926,
2185
+ 927,
2186
+ 928,
2187
+ 929,
2188
+ 930,
2189
+ 931,
2190
+ 932,
2191
+ 933,
2192
+ 934,
2193
+ 935,
2194
+ 936,
2195
+ 937,
2196
+ 938,
2197
+ 939,
2198
+ 940,
2199
+ 941,
2200
+ 942,
2201
+ 943,
2202
+ 944,
2203
+ 945,
2204
+ 946,
2205
+ 947,
2206
+ 948,
2207
+ 949,
2208
+ 950,
2209
+ 951,
2210
+ 952,
2211
+ 953,
2212
+ 954,
2213
+ 955,
2214
+ 956,
2215
+ 957,
2216
+ 958,
2217
+ 959,
2218
+ 960,
2219
+ 961,
2220
+ 962,
2221
+ 963,
2222
+ 964,
2223
+ 965,
2224
+ 966,
2225
+ 967,
2226
+ 968,
2227
+ 969,
2228
+ 970,
2229
+ 971,
2230
+ 972,
2231
+ 973,
2232
+ 974,
2233
+ 975,
2234
+ 976,
2235
+ 977,
2236
+ 978,
2237
+ 979,
2238
+ 980,
2239
+ 981,
2240
+ 982,
2241
+ 983,
2242
+ 984,
2243
+ 985,
2244
+ 986,
2245
+ 987,
2246
+ 988,
2247
+ 989,
2248
+ 990,
2249
+ 991,
2250
+ 992,
2251
+ 993,
2252
+ 994,
2253
+ 995,
2254
+ 996,
2255
+ 997,
2256
+ 998,
2257
+ 999,
2258
+ 1000,
2259
+ 1001,
2260
+ 1002,
2261
+ 1003,
2262
+ 1004,
2263
+ 1005,
2264
+ 1006,
2265
+ 1007,
2266
+ 1008,
2267
+ 1009,
2268
+ 1010,
2269
+ 1011,
2270
+ 1012,
2271
+ 1013,
2272
+ 1014,
2273
+ 1015,
2274
+ 1016,
2275
+ 1017,
2276
+ 1018,
2277
+ 1019,
2278
+ 1020,
2279
+ 1021,
2280
+ 1022,
2281
+ 1023
2282
+ ],
2283
+ "steps": null,
2284
+ "target_fps": 50.0,
2285
+ "valid_shape": [
2286
+ 1024,
2287
+ 199
2288
+ ]
2289
+ },
2290
+ "transition_metrics": {
2291
+ "chunk_length": 60,
2292
+ "enabled": false,
2293
+ "reason": "disabled"
2294
+ }
2295
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/dataset_metrics.json ADDED
@@ -0,0 +1,1046 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "100style_original_filtered_test": {
3
+ "embedding": {
4
+ "diversity_generated": 1.0575839117993924,
5
+ "motion_fid": 0.295132260782035,
6
+ "motion_kid": {
7
+ "mean": 0.00030837465139699205,
8
+ "std": 0.0
9
+ },
10
+ "num_samples": 86
11
+ },
12
+ "num_samples": 86,
13
+ "physical": {
14
+ "body_jerk_mean": {
15
+ "max": 306.9395446777344,
16
+ "mean": 91.73821926116943,
17
+ "min": 14.676246643066406,
18
+ "std": 41.61339123942521
19
+ },
20
+ "contact_sliding_speed": {
21
+ "max": 1.9230101108551025,
22
+ "mean": 0.7933437027318706,
23
+ "min": 0.0035282010212540627,
24
+ "std": 0.39022621340146474
25
+ },
26
+ "foot_ground_error": {
27
+ "max": 0.04590573161840439,
28
+ "mean": 0.013192172581871409,
29
+ "min": 0.0021116272546350956,
30
+ "std": 0.009524469604494283
31
+ },
32
+ "num_samples": 86
33
+ },
34
+ "text": {
35
+ "generated": {
36
+ "matching_distance": {
37
+ "max": 1.3606981039047241,
38
+ "mean": 1.1547354348870211,
39
+ "min": 0.9792652130126953,
40
+ "std": 0.07691267359783768
41
+ },
42
+ "matching_score": 1.1547354459762573,
43
+ "mean_text_rank": 7.109375,
44
+ "median_text_rank": 4.0,
45
+ "r_precision": [
46
+ 0.3125,
47
+ 0.421875,
48
+ 0.453125
49
+ ],
50
+ "retrieval_batch_size": 32,
51
+ "retrieval_num_batches": 2,
52
+ "retrieval_num_samples": 64,
53
+ "retrieval_order_policy": "manifest_order",
54
+ "retrieval_unique_datasets_per_batch": {
55
+ "max": 1.0,
56
+ "mean": 1.0,
57
+ "min": 1.0,
58
+ "std": 0.0
59
+ }
60
+ },
61
+ "reference": {
62
+ "matching_distance": {
63
+ "max": 1.3797343969345093,
64
+ "mean": 1.1303709336491519,
65
+ "min": 0.976786196231842,
66
+ "std": 0.0715656855131584
67
+ },
68
+ "matching_score": 1.130370855331421,
69
+ "mean_text_rank": 6.03125,
70
+ "median_text_rank": 4.0,
71
+ "r_precision": [
72
+ 0.328125,
73
+ 0.453125,
74
+ 0.46875
75
+ ],
76
+ "retrieval_batch_size": 32,
77
+ "retrieval_num_batches": 2,
78
+ "retrieval_num_samples": 64,
79
+ "retrieval_order_policy": "manifest_order",
80
+ "retrieval_unique_datasets_per_batch": {
81
+ "max": 1.0,
82
+ "mean": 1.0,
83
+ "min": 1.0,
84
+ "std": 0.0
85
+ }
86
+ }
87
+ }
88
+ },
89
+ "amass_original_filtered_test": {
90
+ "embedding": {
91
+ "diversity_generated": 1.2720565270928372,
92
+ "motion_fid": 0.3890069629224415,
93
+ "motion_kid": {
94
+ "mean": 7.148098609599529e-05,
95
+ "std": 0.0
96
+ },
97
+ "num_samples": 86
98
+ },
99
+ "num_samples": 86,
100
+ "physical": {
101
+ "body_jerk_mean": {
102
+ "max": 265.1717834472656,
103
+ "mean": 72.45789514585982,
104
+ "min": 11.791740417480469,
105
+ "std": 45.92659264260362
106
+ },
107
+ "contact_sliding_speed": {
108
+ "max": 1.6666642427444458,
109
+ "mean": 0.40420085963045876,
110
+ "min": 0.0,
111
+ "std": 0.4343716464602378
112
+ },
113
+ "foot_ground_error": {
114
+ "max": 0.11975693702697754,
115
+ "mean": 0.026942830335153916,
116
+ "min": 0.0033850527834147215,
117
+ "std": 0.016949850606657043
118
+ },
119
+ "num_samples": 86
120
+ },
121
+ "text": {
122
+ "generated": {
123
+ "matching_distance": {
124
+ "max": 1.5543382167816162,
125
+ "mean": 1.1741622436878294,
126
+ "min": 1.0039037466049194,
127
+ "std": 0.10332746152566549
128
+ },
129
+ "matching_score": 1.1741621494293213,
130
+ "mean_text_rank": 4.84375,
131
+ "median_text_rank": 3.0,
132
+ "r_precision": [
133
+ 0.265625,
134
+ 0.46875,
135
+ 0.5625
136
+ ],
137
+ "retrieval_batch_size": 32,
138
+ "retrieval_num_batches": 2,
139
+ "retrieval_num_samples": 64,
140
+ "retrieval_order_policy": "manifest_order",
141
+ "retrieval_unique_datasets_per_batch": {
142
+ "max": 1.0,
143
+ "mean": 1.0,
144
+ "min": 1.0,
145
+ "std": 0.0
146
+ }
147
+ },
148
+ "reference": {
149
+ "matching_distance": {
150
+ "max": 1.4848475456237793,
151
+ "mean": 1.1559161391369133,
152
+ "min": 0.9887187480926514,
153
+ "std": 0.09615433157926441
154
+ },
155
+ "matching_score": 1.1559160947799683,
156
+ "mean_text_rank": 4.71875,
157
+ "median_text_rank": 3.0,
158
+ "r_precision": [
159
+ 0.3125,
160
+ 0.4375,
161
+ 0.5625
162
+ ],
163
+ "retrieval_batch_size": 32,
164
+ "retrieval_num_batches": 2,
165
+ "retrieval_num_samples": 64,
166
+ "retrieval_order_policy": "manifest_order",
167
+ "retrieval_unique_datasets_per_batch": {
168
+ "max": 1.0,
169
+ "mean": 1.0,
170
+ "min": 1.0,
171
+ "std": 0.0
172
+ }
173
+ }
174
+ }
175
+ },
176
+ "bones_seed_original_filtered_test": {
177
+ "embedding": {
178
+ "diversity_generated": 1.2288775777117131,
179
+ "motion_fid": 0.2935687819807675,
180
+ "motion_kid": {
181
+ "mean": 4.6969598594959194e-05,
182
+ "std": 0.0
183
+ },
184
+ "num_samples": 86
185
+ },
186
+ "num_samples": 86,
187
+ "physical": {
188
+ "body_jerk_mean": {
189
+ "max": 392.7170104980469,
190
+ "mean": 67.97546840268512,
191
+ "min": 14.550009727478027,
192
+ "std": 66.54340159249884
193
+ },
194
+ "contact_sliding_speed": {
195
+ "max": 1.5325274467468262,
196
+ "mean": 0.32937751099099083,
197
+ "min": 0.004231189843267202,
198
+ "std": 0.4110081512966714
199
+ },
200
+ "foot_ground_error": {
201
+ "max": 0.06508437544107437,
202
+ "mean": 0.013941306205260632,
203
+ "min": 0.001276918570511043,
204
+ "std": 0.011664757008804184
205
+ },
206
+ "num_samples": 86
207
+ },
208
+ "text": {
209
+ "generated": {
210
+ "matching_distance": {
211
+ "max": 1.3572005033493042,
212
+ "mean": 1.1043026724527047,
213
+ "min": 0.8937378525733948,
214
+ "std": 0.11051084308488551
215
+ },
216
+ "matching_score": 1.1043027639389038,
217
+ "mean_text_rank": 5.359375,
218
+ "median_text_rank": 2.0,
219
+ "r_precision": [
220
+ 0.46875,
221
+ 0.578125,
222
+ 0.65625
223
+ ],
224
+ "retrieval_batch_size": 32,
225
+ "retrieval_num_batches": 2,
226
+ "retrieval_num_samples": 64,
227
+ "retrieval_order_policy": "manifest_order",
228
+ "retrieval_unique_datasets_per_batch": {
229
+ "max": 1.0,
230
+ "mean": 1.0,
231
+ "min": 1.0,
232
+ "std": 0.0
233
+ }
234
+ },
235
+ "reference": {
236
+ "matching_distance": {
237
+ "max": 1.2490898370742798,
238
+ "mean": 1.0238502039465793,
239
+ "min": 0.8655920624732971,
240
+ "std": 0.08327281471573043
241
+ },
242
+ "matching_score": 1.0238500833511353,
243
+ "mean_text_rank": 2.03125,
244
+ "median_text_rank": 1.0,
245
+ "r_precision": [
246
+ 0.734375,
247
+ 0.828125,
248
+ 0.84375
249
+ ],
250
+ "retrieval_batch_size": 32,
251
+ "retrieval_num_batches": 2,
252
+ "retrieval_num_samples": 64,
253
+ "retrieval_order_policy": "manifest_order",
254
+ "retrieval_unique_datasets_per_batch": {
255
+ "max": 1.0,
256
+ "mean": 1.0,
257
+ "min": 1.0,
258
+ "std": 0.0
259
+ }
260
+ }
261
+ }
262
+ },
263
+ "fitness_original_filtered_test": {
264
+ "embedding": {
265
+ "diversity_generated": 1.2788421887176629,
266
+ "motion_fid": 0.7282577019576995,
267
+ "motion_kid": {
268
+ "mean": 0.0006897810534862714,
269
+ "std": 0.0
270
+ },
271
+ "num_samples": 86
272
+ },
273
+ "num_samples": 86,
274
+ "physical": {
275
+ "body_jerk_mean": {
276
+ "max": 229.07168579101562,
277
+ "mean": 80.73191664939704,
278
+ "min": 22.525007247924805,
279
+ "std": 47.517861751693125
280
+ },
281
+ "contact_sliding_speed": {
282
+ "max": 1.1983479261398315,
283
+ "mean": 0.1302262063917899,
284
+ "min": 0.0,
285
+ "std": 0.19028626905031584
286
+ },
287
+ "foot_ground_error": {
288
+ "max": 0.14234204590320587,
289
+ "mean": 0.07587297533660434,
290
+ "min": 0.017589636147022247,
291
+ "std": 0.021546001091225772
292
+ },
293
+ "num_samples": 86
294
+ },
295
+ "text": {
296
+ "generated": {
297
+ "matching_distance": {
298
+ "max": 1.3765006065368652,
299
+ "mean": 1.192287171995917,
300
+ "min": 1.0051730871200562,
301
+ "std": 0.07750270232156863
302
+ },
303
+ "matching_score": 1.1922872066497803,
304
+ "mean_text_rank": 8.015625,
305
+ "median_text_rank": 5.0,
306
+ "r_precision": [
307
+ 0.359375,
308
+ 0.359375,
309
+ 0.46875
310
+ ],
311
+ "retrieval_batch_size": 32,
312
+ "retrieval_num_batches": 2,
313
+ "retrieval_num_samples": 64,
314
+ "retrieval_order_policy": "manifest_order",
315
+ "retrieval_unique_datasets_per_batch": {
316
+ "max": 1.0,
317
+ "mean": 1.0,
318
+ "min": 1.0,
319
+ "std": 0.0
320
+ }
321
+ },
322
+ "reference": {
323
+ "matching_distance": {
324
+ "max": 1.3280129432678223,
325
+ "mean": 1.1493867725826974,
326
+ "min": 0.9877991080284119,
327
+ "std": 0.08739226669054043
328
+ },
329
+ "matching_score": 1.1493868827819824,
330
+ "mean_text_rank": 9.71875,
331
+ "median_text_rank": 10.0,
332
+ "r_precision": [
333
+ 0.21875,
334
+ 0.21875,
335
+ 0.234375
336
+ ],
337
+ "retrieval_batch_size": 32,
338
+ "retrieval_num_batches": 2,
339
+ "retrieval_num_samples": 64,
340
+ "retrieval_order_policy": "manifest_order",
341
+ "retrieval_unique_datasets_per_batch": {
342
+ "max": 1.0,
343
+ "mean": 1.0,
344
+ "min": 1.0,
345
+ "std": 0.0
346
+ }
347
+ }
348
+ }
349
+ },
350
+ "humanml_original_filtered_test": {
351
+ "embedding": {
352
+ "diversity_generated": 1.2619524338100758,
353
+ "motion_fid": 0.37949637116434376,
354
+ "motion_kid": {
355
+ "mean": 0.0001613574830927611,
356
+ "std": 0.0
357
+ },
358
+ "num_samples": 85
359
+ },
360
+ "num_samples": 85,
361
+ "physical": {
362
+ "body_jerk_mean": {
363
+ "max": 335.8555908203125,
364
+ "mean": 68.35952532151165,
365
+ "min": 16.2313232421875,
366
+ "std": 57.10007770878522
367
+ },
368
+ "contact_sliding_speed": {
369
+ "max": 1.2696173191070557,
370
+ "mean": 0.24481515460099806,
371
+ "min": 0.0,
372
+ "std": 0.26533379757226744
373
+ },
374
+ "foot_ground_error": {
375
+ "max": 0.22183531522750854,
376
+ "mean": 0.07665189722114626,
377
+ "min": 0.005901610944420099,
378
+ "std": 0.03276035910571276
379
+ },
380
+ "num_samples": 85
381
+ },
382
+ "text": {
383
+ "generated": {
384
+ "matching_distance": {
385
+ "max": 1.4400562047958374,
386
+ "mean": 1.119324480084812,
387
+ "min": 0.9264796376228333,
388
+ "std": 0.10692406599540519
389
+ },
390
+ "matching_score": 1.1193243265151978,
391
+ "mean_text_rank": 3.625,
392
+ "median_text_rank": 2.0,
393
+ "r_precision": [
394
+ 0.4375,
395
+ 0.59375,
396
+ 0.65625
397
+ ],
398
+ "retrieval_batch_size": 32,
399
+ "retrieval_num_batches": 2,
400
+ "retrieval_num_samples": 64,
401
+ "retrieval_order_policy": "manifest_order",
402
+ "retrieval_unique_datasets_per_batch": {
403
+ "max": 1.0,
404
+ "mean": 1.0,
405
+ "min": 1.0,
406
+ "std": 0.0
407
+ }
408
+ },
409
+ "reference": {
410
+ "matching_distance": {
411
+ "max": 1.3195350170135498,
412
+ "mean": 1.0584916128831752,
413
+ "min": 0.8765217065811157,
414
+ "std": 0.10127790732110797
415
+ },
416
+ "matching_score": 1.058491587638855,
417
+ "mean_text_rank": 2.71875,
418
+ "median_text_rank": 2.0,
419
+ "r_precision": [
420
+ 0.484375,
421
+ 0.671875,
422
+ 0.75
423
+ ],
424
+ "retrieval_batch_size": 32,
425
+ "retrieval_num_batches": 2,
426
+ "retrieval_num_samples": 64,
427
+ "retrieval_order_policy": "manifest_order",
428
+ "retrieval_unique_datasets_per_batch": {
429
+ "max": 1.0,
430
+ "mean": 1.0,
431
+ "min": 1.0,
432
+ "std": 0.0
433
+ }
434
+ }
435
+ }
436
+ },
437
+ "idea400_original_filtered_test": {
438
+ "embedding": {
439
+ "diversity_generated": 1.1603107092959224,
440
+ "motion_fid": 0.3089720542673662,
441
+ "motion_kid": {
442
+ "mean": 0.00024784405716316016,
443
+ "std": 0.0
444
+ },
445
+ "num_samples": 85
446
+ },
447
+ "num_samples": 85,
448
+ "physical": {
449
+ "body_jerk_mean": {
450
+ "max": 109.74834442138672,
451
+ "mean": 46.92937700608197,
452
+ "min": 12.396353721618652,
453
+ "std": 20.873435730428675
454
+ },
455
+ "contact_sliding_speed": {
456
+ "max": 0.8785831928253174,
457
+ "mean": 0.14769231055808418,
458
+ "min": 0.0,
459
+ "std": 0.24178951340825944
460
+ },
461
+ "foot_ground_error": {
462
+ "max": 0.2917259931564331,
463
+ "mean": 0.0424500749556019,
464
+ "min": 0.007580623496323824,
465
+ "std": 0.03828018100277618
466
+ },
467
+ "num_samples": 85
468
+ },
469
+ "text": {
470
+ "generated": {
471
+ "matching_distance": {
472
+ "max": 1.4350745677947998,
473
+ "mean": 1.1293431899126838,
474
+ "min": 0.9147061705589294,
475
+ "std": 0.10790011773020315
476
+ },
477
+ "matching_score": 1.1293431520462036,
478
+ "mean_text_rank": 7.203125,
479
+ "median_text_rank": 5.0,
480
+ "r_precision": [
481
+ 0.1875,
482
+ 0.34375,
483
+ 0.4375
484
+ ],
485
+ "retrieval_batch_size": 32,
486
+ "retrieval_num_batches": 2,
487
+ "retrieval_num_samples": 64,
488
+ "retrieval_order_policy": "manifest_order",
489
+ "retrieval_unique_datasets_per_batch": {
490
+ "max": 1.0,
491
+ "mean": 1.0,
492
+ "min": 1.0,
493
+ "std": 0.0
494
+ }
495
+ },
496
+ "reference": {
497
+ "matching_distance": {
498
+ "max": 1.3933959007263184,
499
+ "mean": 1.0893783358966602,
500
+ "min": 0.8957217931747437,
501
+ "std": 0.09054021801661577
502
+ },
503
+ "matching_score": 1.0893783569335938,
504
+ "mean_text_rank": 6.359375,
505
+ "median_text_rank": 4.0,
506
+ "r_precision": [
507
+ 0.21875,
508
+ 0.359375,
509
+ 0.46875
510
+ ],
511
+ "retrieval_batch_size": 32,
512
+ "retrieval_num_batches": 2,
513
+ "retrieval_num_samples": 64,
514
+ "retrieval_order_policy": "manifest_order",
515
+ "retrieval_unique_datasets_per_batch": {
516
+ "max": 1.0,
517
+ "mean": 1.0,
518
+ "min": 1.0,
519
+ "std": 0.0
520
+ }
521
+ }
522
+ }
523
+ },
524
+ "lafan1_original_filtered_test": {
525
+ "embedding": {
526
+ "diversity_generated": 1.0652331521765446,
527
+ "motion_fid": 0.4122437081306951,
528
+ "motion_kid": {
529
+ "mean": 0.00044010696667928784,
530
+ "std": 0.0
531
+ },
532
+ "num_samples": 85
533
+ },
534
+ "num_samples": 85,
535
+ "physical": {
536
+ "body_jerk_mean": {
537
+ "max": 314.1806945800781,
538
+ "mean": 139.9726458381204,
539
+ "min": 31.704696655273438,
540
+ "std": 71.70045395223674
541
+ },
542
+ "contact_sliding_speed": {
543
+ "max": 2.5723462104797363,
544
+ "mean": 1.074749160612769,
545
+ "min": 0.008806232362985611,
546
+ "std": 0.6605997776393984
547
+ },
548
+ "foot_ground_error": {
549
+ "max": 0.11140980571508408,
550
+ "mean": 0.021036932756781906,
551
+ "min": 0.00042374650365673006,
552
+ "std": 0.020322981130113965
553
+ },
554
+ "num_samples": 85
555
+ },
556
+ "text": {
557
+ "generated": {
558
+ "matching_distance": {
559
+ "max": 1.44863760471344,
560
+ "mean": 1.1349114824743831,
561
+ "min": 0.935748815536499,
562
+ "std": 0.10606785314295356
563
+ },
564
+ "matching_score": 1.1349115371704102,
565
+ "mean_text_rank": 8.84375,
566
+ "median_text_rank": 7.5,
567
+ "r_precision": [
568
+ 0.328125,
569
+ 0.359375,
570
+ 0.421875
571
+ ],
572
+ "retrieval_batch_size": 32,
573
+ "retrieval_num_batches": 2,
574
+ "retrieval_num_samples": 64,
575
+ "retrieval_order_policy": "manifest_order",
576
+ "retrieval_unique_datasets_per_batch": {
577
+ "max": 1.0,
578
+ "mean": 1.0,
579
+ "min": 1.0,
580
+ "std": 0.0
581
+ }
582
+ },
583
+ "reference": {
584
+ "matching_distance": {
585
+ "max": 1.4183552265167236,
586
+ "mean": 1.1475334244615891,
587
+ "min": 0.9194038510322571,
588
+ "std": 0.09896612738362433
589
+ },
590
+ "matching_score": 1.1475334167480469,
591
+ "mean_text_rank": 8.109375,
592
+ "median_text_rank": 5.0,
593
+ "r_precision": [
594
+ 0.359375,
595
+ 0.359375,
596
+ 0.375
597
+ ],
598
+ "retrieval_batch_size": 32,
599
+ "retrieval_num_batches": 2,
600
+ "retrieval_num_samples": 64,
601
+ "retrieval_order_policy": "manifest_order",
602
+ "retrieval_unique_datasets_per_batch": {
603
+ "max": 1.0,
604
+ "mean": 1.0,
605
+ "min": 1.0,
606
+ "std": 0.0
607
+ }
608
+ }
609
+ }
610
+ },
611
+ "motiongv_original_filtered_test": {
612
+ "embedding": {
613
+ "diversity_generated": 1.3252344922258321,
614
+ "motion_fid": 0.3607463810043918,
615
+ "motion_kid": {
616
+ "mean": 1.8889422511048082e-05,
617
+ "std": 0.0
618
+ },
619
+ "num_samples": 85
620
+ },
621
+ "num_samples": 85,
622
+ "physical": {
623
+ "body_jerk_mean": {
624
+ "max": 394.6324157714844,
625
+ "mean": 63.257873658572926,
626
+ "min": 10.021343231201172,
627
+ "std": 58.67747854537708
628
+ },
629
+ "contact_sliding_speed": {
630
+ "max": 0.7465254664421082,
631
+ "mean": 0.12326823623905725,
632
+ "min": 0.003070311388000846,
633
+ "std": 0.17676128678585645
634
+ },
635
+ "foot_ground_error": {
636
+ "max": 0.1117943823337555,
637
+ "mean": 0.07272094489239594,
638
+ "min": 0.012083700858056545,
639
+ "std": 0.01693819111508039
640
+ },
641
+ "num_samples": 85
642
+ },
643
+ "text": {
644
+ "generated": {
645
+ "matching_distance": {
646
+ "max": 1.4481498003005981,
647
+ "mean": 1.1457594983717974,
648
+ "min": 0.8835017085075378,
649
+ "std": 0.10666800116978434
650
+ },
651
+ "matching_score": 1.1457593441009521,
652
+ "mean_text_rank": 3.875,
653
+ "median_text_rank": 2.0,
654
+ "r_precision": [
655
+ 0.46875,
656
+ 0.59375,
657
+ 0.703125
658
+ ],
659
+ "retrieval_batch_size": 32,
660
+ "retrieval_num_batches": 2,
661
+ "retrieval_num_samples": 64,
662
+ "retrieval_order_policy": "manifest_order",
663
+ "retrieval_unique_datasets_per_batch": {
664
+ "max": 1.0,
665
+ "mean": 1.0,
666
+ "min": 1.0,
667
+ "std": 0.0
668
+ }
669
+ },
670
+ "reference": {
671
+ "matching_distance": {
672
+ "max": 1.3658052682876587,
673
+ "mean": 1.1116458219640395,
674
+ "min": 0.8830475211143494,
675
+ "std": 0.08860567172404131
676
+ },
677
+ "matching_score": 1.1116458177566528,
678
+ "mean_text_rank": 2.40625,
679
+ "median_text_rank": 1.0,
680
+ "r_precision": [
681
+ 0.609375,
682
+ 0.703125,
683
+ 0.828125
684
+ ],
685
+ "retrieval_batch_size": 32,
686
+ "retrieval_num_batches": 2,
687
+ "retrieval_num_samples": 64,
688
+ "retrieval_order_policy": "manifest_order",
689
+ "retrieval_unique_datasets_per_batch": {
690
+ "max": 1.0,
691
+ "mean": 1.0,
692
+ "min": 1.0,
693
+ "std": 0.0
694
+ }
695
+ }
696
+ }
697
+ },
698
+ "motionllama_original_filtered_test": {
699
+ "embedding": {
700
+ "diversity_generated": 1.243081346203865,
701
+ "motion_fid": 0.41212859468716423,
702
+ "motion_kid": {
703
+ "mean": 0.00039907121682780655,
704
+ "std": 0.0
705
+ },
706
+ "num_samples": 85
707
+ },
708
+ "num_samples": 85,
709
+ "physical": {
710
+ "body_jerk_mean": {
711
+ "max": 149.49559020996094,
712
+ "mean": 51.222771689471074,
713
+ "min": 11.60899543762207,
714
+ "std": 21.585048177328567
715
+ },
716
+ "contact_sliding_speed": {
717
+ "max": 0.3608182966709137,
718
+ "mean": 0.07486999163623242,
719
+ "min": 0.004085029475390911,
720
+ "std": 0.07851552799603706
721
+ },
722
+ "foot_ground_error": {
723
+ "max": 0.10084258019924164,
724
+ "mean": 0.06930571288308676,
725
+ "min": 0.012543153017759323,
726
+ "std": 0.017269924728060666
727
+ },
728
+ "num_samples": 85
729
+ },
730
+ "text": {
731
+ "generated": {
732
+ "matching_distance": {
733
+ "max": 1.3583039045333862,
734
+ "mean": 1.129518364457523,
735
+ "min": 0.907444417476654,
736
+ "std": 0.11405018318730081
737
+ },
738
+ "matching_score": 1.1295183897018433,
739
+ "mean_text_rank": 9.53125,
740
+ "median_text_rank": 7.0,
741
+ "r_precision": [
742
+ 0.171875,
743
+ 0.25,
744
+ 0.28125
745
+ ],
746
+ "retrieval_batch_size": 32,
747
+ "retrieval_num_batches": 2,
748
+ "retrieval_num_samples": 64,
749
+ "retrieval_order_policy": "manifest_order",
750
+ "retrieval_unique_datasets_per_batch": {
751
+ "max": 1.0,
752
+ "mean": 1.0,
753
+ "min": 1.0,
754
+ "std": 0.0
755
+ }
756
+ },
757
+ "reference": {
758
+ "matching_distance": {
759
+ "max": 1.2308001518249512,
760
+ "mean": 1.0292582687209635,
761
+ "min": 0.8811001777648926,
762
+ "std": 0.0877342679085498
763
+ },
764
+ "matching_score": 1.029258131980896,
765
+ "mean_text_rank": 6.921875,
766
+ "median_text_rank": 4.0,
767
+ "r_precision": [
768
+ 0.203125,
769
+ 0.3125,
770
+ 0.46875
771
+ ],
772
+ "retrieval_batch_size": 32,
773
+ "retrieval_num_batches": 2,
774
+ "retrieval_num_samples": 64,
775
+ "retrieval_order_policy": "manifest_order",
776
+ "retrieval_unique_datasets_per_batch": {
777
+ "max": 1.0,
778
+ "mean": 1.0,
779
+ "min": 1.0,
780
+ "std": 0.0
781
+ }
782
+ }
783
+ }
784
+ },
785
+ "omomo_original_filtered_test": {
786
+ "embedding": {
787
+ "diversity_generated": 1.225772022315077,
788
+ "motion_fid": 0.4057685475174302,
789
+ "motion_kid": {
790
+ "mean": 0.00022337189809062963,
791
+ "std": 0.0
792
+ },
793
+ "num_samples": 85
794
+ },
795
+ "num_samples": 85,
796
+ "physical": {
797
+ "body_jerk_mean": {
798
+ "max": 170.46762084960938,
799
+ "mean": 58.10617489534266,
800
+ "min": 16.497365951538086,
801
+ "std": 24.335758018848132
802
+ },
803
+ "contact_sliding_speed": {
804
+ "max": 1.117565631866455,
805
+ "mean": 0.3834191459504997,
806
+ "min": 0.005078582558780909,
807
+ "std": 0.3170930006320161
808
+ },
809
+ "foot_ground_error": {
810
+ "max": 0.1168566644191742,
811
+ "mean": 0.016199057431453293,
812
+ "min": 0.002339268336072564,
813
+ "std": 0.013331415234509801
814
+ },
815
+ "num_samples": 85
816
+ },
817
+ "text": {
818
+ "generated": {
819
+ "matching_distance": {
820
+ "max": 1.4282751083374023,
821
+ "mean": 1.1995593414587133,
822
+ "min": 0.970937192440033,
823
+ "std": 0.09437652461416668
824
+ },
825
+ "matching_score": 1.1995593309402466,
826
+ "mean_text_rank": 9.390625,
827
+ "median_text_rank": 7.0,
828
+ "r_precision": [
829
+ 0.234375,
830
+ 0.328125,
831
+ 0.34375
832
+ ],
833
+ "retrieval_batch_size": 32,
834
+ "retrieval_num_batches": 2,
835
+ "retrieval_num_samples": 64,
836
+ "retrieval_order_policy": "manifest_order",
837
+ "retrieval_unique_datasets_per_batch": {
838
+ "max": 1.0,
839
+ "mean": 1.0,
840
+ "min": 1.0,
841
+ "std": 0.0
842
+ }
843
+ },
844
+ "reference": {
845
+ "matching_distance": {
846
+ "max": 1.3254883289337158,
847
+ "mean": 1.1843213705455555,
848
+ "min": 0.958416759967804,
849
+ "std": 0.07360095182352315
850
+ },
851
+ "matching_score": 1.1843212842941284,
852
+ "mean_text_rank": 7.484375,
853
+ "median_text_rank": 5.5,
854
+ "r_precision": [
855
+ 0.140625,
856
+ 0.328125,
857
+ 0.40625
858
+ ],
859
+ "retrieval_batch_size": 32,
860
+ "retrieval_num_batches": 2,
861
+ "retrieval_num_samples": 64,
862
+ "retrieval_order_policy": "manifest_order",
863
+ "retrieval_unique_datasets_per_batch": {
864
+ "max": 1.0,
865
+ "mean": 1.0,
866
+ "min": 1.0,
867
+ "std": 0.0
868
+ }
869
+ }
870
+ }
871
+ },
872
+ "permo_original_filtered_test": {
873
+ "embedding": {
874
+ "diversity_generated": 1.2086325566904998,
875
+ "motion_fid": 0.4148669497066708,
876
+ "motion_kid": {
877
+ "mean": 0.0003991512469281311,
878
+ "std": 0.0
879
+ },
880
+ "num_samples": 85
881
+ },
882
+ "num_samples": 85,
883
+ "physical": {
884
+ "body_jerk_mean": {
885
+ "max": 240.15951538085938,
886
+ "mean": 89.35240687202005,
887
+ "min": 24.114988327026367,
888
+ "std": 42.25908942517282
889
+ },
890
+ "contact_sliding_speed": {
891
+ "max": 1.3213852643966675,
892
+ "mean": 0.38401777897687517,
893
+ "min": 0.0,
894
+ "std": 0.3190618849745317
895
+ },
896
+ "foot_ground_error": {
897
+ "max": 0.2258841097354889,
898
+ "mean": 0.036138936230803234,
899
+ "min": 0.008152862079441547,
900
+ "std": 0.02827083711142457
901
+ },
902
+ "num_samples": 85
903
+ },
904
+ "text": {
905
+ "generated": {
906
+ "matching_distance": {
907
+ "max": 1.412964940071106,
908
+ "mean": 1.104096428085776,
909
+ "min": 0.8289653062820435,
910
+ "std": 0.12450774868152055
911
+ },
912
+ "matching_score": 1.1040964126586914,
913
+ "mean_text_rank": 5.796875,
914
+ "median_text_rank": 4.0,
915
+ "r_precision": [
916
+ 0.15625,
917
+ 0.359375,
918
+ 0.4375
919
+ ],
920
+ "retrieval_batch_size": 32,
921
+ "retrieval_num_batches": 2,
922
+ "retrieval_num_samples": 64,
923
+ "retrieval_order_policy": "manifest_order",
924
+ "retrieval_unique_datasets_per_batch": {
925
+ "max": 1.0,
926
+ "mean": 1.0,
927
+ "min": 1.0,
928
+ "std": 0.0
929
+ }
930
+ },
931
+ "reference": {
932
+ "matching_distance": {
933
+ "max": 1.2057435512542725,
934
+ "mean": 1.0011776215889874,
935
+ "min": 0.8405618071556091,
936
+ "std": 0.0876302018200968
937
+ },
938
+ "matching_score": 1.0011776685714722,
939
+ "mean_text_rank": 4.84375,
940
+ "median_text_rank": 4.0,
941
+ "r_precision": [
942
+ 0.171875,
943
+ 0.328125,
944
+ 0.46875
945
+ ],
946
+ "retrieval_batch_size": 32,
947
+ "retrieval_num_batches": 2,
948
+ "retrieval_num_samples": 64,
949
+ "retrieval_order_policy": "manifest_order",
950
+ "retrieval_unique_datasets_per_batch": {
951
+ "max": 1.0,
952
+ "mean": 1.0,
953
+ "min": 1.0,
954
+ "std": 0.0
955
+ }
956
+ }
957
+ }
958
+ },
959
+ "snapmogen_original_filtered_test": {
960
+ "embedding": {
961
+ "diversity_generated": 1.1717517212600463,
962
+ "motion_fid": 0.3023945940485344,
963
+ "motion_kid": {
964
+ "mean": 0.0002671196633574624,
965
+ "std": 0.0
966
+ },
967
+ "num_samples": 85
968
+ },
969
+ "num_samples": 85,
970
+ "physical": {
971
+ "body_jerk_mean": {
972
+ "max": 183.62680053710938,
973
+ "mean": 57.84897658404182,
974
+ "min": 16.209308624267578,
975
+ "std": 38.09984754580235
976
+ },
977
+ "contact_sliding_speed": {
978
+ "max": 1.0688523054122925,
979
+ "mean": 0.1868419358452015,
980
+ "min": 0.0041826097294688225,
981
+ "std": 0.20431294637395492
982
+ },
983
+ "foot_ground_error": {
984
+ "max": 0.14229609072208405,
985
+ "mean": 0.034169711305431144,
986
+ "min": 0.0016218408709391952,
987
+ "std": 0.02866694918161823
988
+ },
989
+ "num_samples": 85
990
+ },
991
+ "text": {
992
+ "generated": {
993
+ "matching_distance": {
994
+ "max": 1.3117051124572754,
995
+ "mean": 1.0969315627041984,
996
+ "min": 0.9296107888221741,
997
+ "std": 0.08918784110822944
998
+ },
999
+ "matching_score": 1.0969316959381104,
1000
+ "mean_text_rank": 4.265625,
1001
+ "median_text_rank": 3.0,
1002
+ "r_precision": [
1003
+ 0.359375,
1004
+ 0.46875,
1005
+ 0.65625
1006
+ ],
1007
+ "retrieval_batch_size": 32,
1008
+ "retrieval_num_batches": 2,
1009
+ "retrieval_num_samples": 64,
1010
+ "retrieval_order_policy": "manifest_order",
1011
+ "retrieval_unique_datasets_per_batch": {
1012
+ "max": 1.0,
1013
+ "mean": 1.0,
1014
+ "min": 1.0,
1015
+ "std": 0.0
1016
+ }
1017
+ },
1018
+ "reference": {
1019
+ "matching_distance": {
1020
+ "max": 1.2125790119171143,
1021
+ "mean": 1.0510956680073458,
1022
+ "min": 0.8823143839836121,
1023
+ "std": 0.08172085751280951
1024
+ },
1025
+ "matching_score": 1.051095724105835,
1026
+ "mean_text_rank": 3.140625,
1027
+ "median_text_rank": 2.0,
1028
+ "r_precision": [
1029
+ 0.46875,
1030
+ 0.671875,
1031
+ 0.765625
1032
+ ],
1033
+ "retrieval_batch_size": 32,
1034
+ "retrieval_num_batches": 2,
1035
+ "retrieval_num_samples": 64,
1036
+ "retrieval_order_policy": "manifest_order",
1037
+ "retrieval_unique_datasets_per_batch": {
1038
+ "max": 1.0,
1039
+ "mean": 1.0,
1040
+ "min": 1.0,
1041
+ "std": 0.0
1042
+ }
1043
+ }
1044
+ }
1045
+ }
1046
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/embedding_metrics.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "diversity_generated": 1.3800554760633976,
3
+ "embedding_frames": 120,
4
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
5
+ "matching_score_generated": 1.1404733657836914,
6
+ "matching_score_reference": 1.0944490432739258,
7
+ "motion_fid": 0.10967632126673776,
8
+ "motion_key": "body_pos_local",
9
+ "motion_kid": {
10
+ "mean": 7.427152145522697e-05,
11
+ "std": 0.0
12
+ },
13
+ "multimodality": null,
14
+ "num_samples": 1024,
15
+ "r_precision_generated": [
16
+ 0.5869140625,
17
+ 0.7333984375,
18
+ 0.818359375
19
+ ],
20
+ "r_precision_reference": [
21
+ 0.6611328125,
22
+ 0.82421875,
23
+ 0.8916015625
24
+ ],
25
+ "text_cache_dir": "${OMG_REPO_ROOT}/outputs/evaluator_text_cache/filtered_original_g1_tmr_bodypos_60f",
26
+ "text_encoder_model": "${OMG_REPO_ROOT}/models/t5-3b-local",
27
+ "text_generated": {
28
+ "matching_distance": {
29
+ "max": 1.5543382167816162,
30
+ "mean": 1.1404733364470303,
31
+ "min": 0.8289653062820435,
32
+ "std": 0.10747345553020096
33
+ },
34
+ "matching_score": 1.1404733657836914,
35
+ "mean_text_rank": 2.576171875,
36
+ "median_text_rank": 1.0,
37
+ "r_precision": [
38
+ 0.5869140625,
39
+ 0.7333984375,
40
+ 0.818359375
41
+ ],
42
+ "retrieval_batch_size": 32,
43
+ "retrieval_num_batches": 32,
44
+ "retrieval_num_samples": 1024,
45
+ "retrieval_order_policy": "stratified_by_dataset",
46
+ "retrieval_unique_datasets_per_batch": {
47
+ "max": 12.0,
48
+ "mean": 11.9375,
49
+ "min": 10.0,
50
+ "std": 0.34798527267687634
51
+ }
52
+ },
53
+ "text_reference": {
54
+ "matching_distance": {
55
+ "max": 1.4848475456237793,
56
+ "mean": 1.0944489734247327,
57
+ "min": 0.8405618071556091,
58
+ "std": 0.10523510496847523
59
+ },
60
+ "matching_score": 1.0944490432739258,
61
+ "mean_text_rank": 1.888671875,
62
+ "median_text_rank": 1.0,
63
+ "r_precision": [
64
+ 0.6611328125,
65
+ 0.82421875,
66
+ 0.8916015625
67
+ ],
68
+ "retrieval_batch_size": 32,
69
+ "retrieval_num_batches": 32,
70
+ "retrieval_num_samples": 1024,
71
+ "retrieval_order_policy": "stratified_by_dataset",
72
+ "retrieval_unique_datasets_per_batch": {
73
+ "max": 12.0,
74
+ "mean": 11.9375,
75
+ "min": 10.0,
76
+ "std": 0.34798527267687634
77
+ }
78
+ },
79
+ "text_retrieval_batch_size": 32
80
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/merge_summary.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_samples/mixed_modalities_all_v1/text_test_1024.jsonl",
3
+ "output_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/generated_qpos_raw.npz",
4
+ "num_samples": 1024,
5
+ "num_frames": 120,
6
+ "num_shards": 8,
7
+ "shard_artifacts": [
8
+ "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/generated_qpos.npz",
9
+ "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_001/generated_qpos.npz",
10
+ "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_002/generated_qpos.npz",
11
+ "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_003/generated_qpos.npz",
12
+ "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_004/generated_qpos.npz",
13
+ "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_005/generated_qpos.npz",
14
+ "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_006/generated_qpos.npz",
15
+ "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_007/generated_qpos.npz"
16
+ ]
17
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/metrics.json ADDED
@@ -0,0 +1,2298 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "benchmark": {
3
+ "benchmark": "artifact_text",
4
+ "dataset_metrics": {
5
+ "100style_original_filtered_test": {
6
+ "embedding": {
7
+ "diversity_generated": 1.0575839117993924,
8
+ "motion_fid": 0.295132260782035,
9
+ "motion_kid": {
10
+ "mean": 0.00030837465139699205,
11
+ "std": 0.0
12
+ },
13
+ "num_samples": 86
14
+ },
15
+ "num_samples": 86,
16
+ "physical": {
17
+ "body_jerk_mean": {
18
+ "max": 306.9395446777344,
19
+ "mean": 91.73821926116943,
20
+ "min": 14.676246643066406,
21
+ "std": 41.61339123942521
22
+ },
23
+ "contact_sliding_speed": {
24
+ "max": 1.9230101108551025,
25
+ "mean": 0.7933437027318706,
26
+ "min": 0.0035282010212540627,
27
+ "std": 0.39022621340146474
28
+ },
29
+ "foot_ground_error": {
30
+ "max": 0.04590573161840439,
31
+ "mean": 0.013192172581871409,
32
+ "min": 0.0021116272546350956,
33
+ "std": 0.009524469604494283
34
+ },
35
+ "num_samples": 86
36
+ },
37
+ "text": {
38
+ "generated": {
39
+ "matching_distance": {
40
+ "max": 1.3606981039047241,
41
+ "mean": 1.1547354348870211,
42
+ "min": 0.9792652130126953,
43
+ "std": 0.07691267359783768
44
+ },
45
+ "matching_score": 1.1547354459762573,
46
+ "mean_text_rank": 7.109375,
47
+ "median_text_rank": 4.0,
48
+ "r_precision": [
49
+ 0.3125,
50
+ 0.421875,
51
+ 0.453125
52
+ ],
53
+ "retrieval_batch_size": 32,
54
+ "retrieval_num_batches": 2,
55
+ "retrieval_num_samples": 64,
56
+ "retrieval_order_policy": "manifest_order",
57
+ "retrieval_unique_datasets_per_batch": {
58
+ "max": 1.0,
59
+ "mean": 1.0,
60
+ "min": 1.0,
61
+ "std": 0.0
62
+ }
63
+ },
64
+ "reference": {
65
+ "matching_distance": {
66
+ "max": 1.3797343969345093,
67
+ "mean": 1.1303709336491519,
68
+ "min": 0.976786196231842,
69
+ "std": 0.0715656855131584
70
+ },
71
+ "matching_score": 1.130370855331421,
72
+ "mean_text_rank": 6.03125,
73
+ "median_text_rank": 4.0,
74
+ "r_precision": [
75
+ 0.328125,
76
+ 0.453125,
77
+ 0.46875
78
+ ],
79
+ "retrieval_batch_size": 32,
80
+ "retrieval_num_batches": 2,
81
+ "retrieval_num_samples": 64,
82
+ "retrieval_order_policy": "manifest_order",
83
+ "retrieval_unique_datasets_per_batch": {
84
+ "max": 1.0,
85
+ "mean": 1.0,
86
+ "min": 1.0,
87
+ "std": 0.0
88
+ }
89
+ }
90
+ }
91
+ },
92
+ "amass_original_filtered_test": {
93
+ "embedding": {
94
+ "diversity_generated": 1.2720565270928372,
95
+ "motion_fid": 0.3890069629224415,
96
+ "motion_kid": {
97
+ "mean": 7.148098609599529e-05,
98
+ "std": 0.0
99
+ },
100
+ "num_samples": 86
101
+ },
102
+ "num_samples": 86,
103
+ "physical": {
104
+ "body_jerk_mean": {
105
+ "max": 265.1717834472656,
106
+ "mean": 72.45789514585982,
107
+ "min": 11.791740417480469,
108
+ "std": 45.92659264260362
109
+ },
110
+ "contact_sliding_speed": {
111
+ "max": 1.6666642427444458,
112
+ "mean": 0.40420085963045876,
113
+ "min": 0.0,
114
+ "std": 0.4343716464602378
115
+ },
116
+ "foot_ground_error": {
117
+ "max": 0.11975693702697754,
118
+ "mean": 0.026942830335153916,
119
+ "min": 0.0033850527834147215,
120
+ "std": 0.016949850606657043
121
+ },
122
+ "num_samples": 86
123
+ },
124
+ "text": {
125
+ "generated": {
126
+ "matching_distance": {
127
+ "max": 1.5543382167816162,
128
+ "mean": 1.1741622436878294,
129
+ "min": 1.0039037466049194,
130
+ "std": 0.10332746152566549
131
+ },
132
+ "matching_score": 1.1741621494293213,
133
+ "mean_text_rank": 4.84375,
134
+ "median_text_rank": 3.0,
135
+ "r_precision": [
136
+ 0.265625,
137
+ 0.46875,
138
+ 0.5625
139
+ ],
140
+ "retrieval_batch_size": 32,
141
+ "retrieval_num_batches": 2,
142
+ "retrieval_num_samples": 64,
143
+ "retrieval_order_policy": "manifest_order",
144
+ "retrieval_unique_datasets_per_batch": {
145
+ "max": 1.0,
146
+ "mean": 1.0,
147
+ "min": 1.0,
148
+ "std": 0.0
149
+ }
150
+ },
151
+ "reference": {
152
+ "matching_distance": {
153
+ "max": 1.4848475456237793,
154
+ "mean": 1.1559161391369133,
155
+ "min": 0.9887187480926514,
156
+ "std": 0.09615433157926441
157
+ },
158
+ "matching_score": 1.1559160947799683,
159
+ "mean_text_rank": 4.71875,
160
+ "median_text_rank": 3.0,
161
+ "r_precision": [
162
+ 0.3125,
163
+ 0.4375,
164
+ 0.5625
165
+ ],
166
+ "retrieval_batch_size": 32,
167
+ "retrieval_num_batches": 2,
168
+ "retrieval_num_samples": 64,
169
+ "retrieval_order_policy": "manifest_order",
170
+ "retrieval_unique_datasets_per_batch": {
171
+ "max": 1.0,
172
+ "mean": 1.0,
173
+ "min": 1.0,
174
+ "std": 0.0
175
+ }
176
+ }
177
+ }
178
+ },
179
+ "bones_seed_original_filtered_test": {
180
+ "embedding": {
181
+ "diversity_generated": 1.2288775777117131,
182
+ "motion_fid": 0.2935687819807675,
183
+ "motion_kid": {
184
+ "mean": 4.6969598594959194e-05,
185
+ "std": 0.0
186
+ },
187
+ "num_samples": 86
188
+ },
189
+ "num_samples": 86,
190
+ "physical": {
191
+ "body_jerk_mean": {
192
+ "max": 392.7170104980469,
193
+ "mean": 67.97546840268512,
194
+ "min": 14.550009727478027,
195
+ "std": 66.54340159249884
196
+ },
197
+ "contact_sliding_speed": {
198
+ "max": 1.5325274467468262,
199
+ "mean": 0.32937751099099083,
200
+ "min": 0.004231189843267202,
201
+ "std": 0.4110081512966714
202
+ },
203
+ "foot_ground_error": {
204
+ "max": 0.06508437544107437,
205
+ "mean": 0.013941306205260632,
206
+ "min": 0.001276918570511043,
207
+ "std": 0.011664757008804184
208
+ },
209
+ "num_samples": 86
210
+ },
211
+ "text": {
212
+ "generated": {
213
+ "matching_distance": {
214
+ "max": 1.3572005033493042,
215
+ "mean": 1.1043026724527047,
216
+ "min": 0.8937378525733948,
217
+ "std": 0.11051084308488551
218
+ },
219
+ "matching_score": 1.1043027639389038,
220
+ "mean_text_rank": 5.359375,
221
+ "median_text_rank": 2.0,
222
+ "r_precision": [
223
+ 0.46875,
224
+ 0.578125,
225
+ 0.65625
226
+ ],
227
+ "retrieval_batch_size": 32,
228
+ "retrieval_num_batches": 2,
229
+ "retrieval_num_samples": 64,
230
+ "retrieval_order_policy": "manifest_order",
231
+ "retrieval_unique_datasets_per_batch": {
232
+ "max": 1.0,
233
+ "mean": 1.0,
234
+ "min": 1.0,
235
+ "std": 0.0
236
+ }
237
+ },
238
+ "reference": {
239
+ "matching_distance": {
240
+ "max": 1.2490898370742798,
241
+ "mean": 1.0238502039465793,
242
+ "min": 0.8655920624732971,
243
+ "std": 0.08327281471573043
244
+ },
245
+ "matching_score": 1.0238500833511353,
246
+ "mean_text_rank": 2.03125,
247
+ "median_text_rank": 1.0,
248
+ "r_precision": [
249
+ 0.734375,
250
+ 0.828125,
251
+ 0.84375
252
+ ],
253
+ "retrieval_batch_size": 32,
254
+ "retrieval_num_batches": 2,
255
+ "retrieval_num_samples": 64,
256
+ "retrieval_order_policy": "manifest_order",
257
+ "retrieval_unique_datasets_per_batch": {
258
+ "max": 1.0,
259
+ "mean": 1.0,
260
+ "min": 1.0,
261
+ "std": 0.0
262
+ }
263
+ }
264
+ }
265
+ },
266
+ "fitness_original_filtered_test": {
267
+ "embedding": {
268
+ "diversity_generated": 1.2788421887176629,
269
+ "motion_fid": 0.7282577019576995,
270
+ "motion_kid": {
271
+ "mean": 0.0006897810534862714,
272
+ "std": 0.0
273
+ },
274
+ "num_samples": 86
275
+ },
276
+ "num_samples": 86,
277
+ "physical": {
278
+ "body_jerk_mean": {
279
+ "max": 229.07168579101562,
280
+ "mean": 80.73191664939704,
281
+ "min": 22.525007247924805,
282
+ "std": 47.517861751693125
283
+ },
284
+ "contact_sliding_speed": {
285
+ "max": 1.1983479261398315,
286
+ "mean": 0.1302262063917899,
287
+ "min": 0.0,
288
+ "std": 0.19028626905031584
289
+ },
290
+ "foot_ground_error": {
291
+ "max": 0.14234204590320587,
292
+ "mean": 0.07587297533660434,
293
+ "min": 0.017589636147022247,
294
+ "std": 0.021546001091225772
295
+ },
296
+ "num_samples": 86
297
+ },
298
+ "text": {
299
+ "generated": {
300
+ "matching_distance": {
301
+ "max": 1.3765006065368652,
302
+ "mean": 1.192287171995917,
303
+ "min": 1.0051730871200562,
304
+ "std": 0.07750270232156863
305
+ },
306
+ "matching_score": 1.1922872066497803,
307
+ "mean_text_rank": 8.015625,
308
+ "median_text_rank": 5.0,
309
+ "r_precision": [
310
+ 0.359375,
311
+ 0.359375,
312
+ 0.46875
313
+ ],
314
+ "retrieval_batch_size": 32,
315
+ "retrieval_num_batches": 2,
316
+ "retrieval_num_samples": 64,
317
+ "retrieval_order_policy": "manifest_order",
318
+ "retrieval_unique_datasets_per_batch": {
319
+ "max": 1.0,
320
+ "mean": 1.0,
321
+ "min": 1.0,
322
+ "std": 0.0
323
+ }
324
+ },
325
+ "reference": {
326
+ "matching_distance": {
327
+ "max": 1.3280129432678223,
328
+ "mean": 1.1493867725826974,
329
+ "min": 0.9877991080284119,
330
+ "std": 0.08739226669054043
331
+ },
332
+ "matching_score": 1.1493868827819824,
333
+ "mean_text_rank": 9.71875,
334
+ "median_text_rank": 10.0,
335
+ "r_precision": [
336
+ 0.21875,
337
+ 0.21875,
338
+ 0.234375
339
+ ],
340
+ "retrieval_batch_size": 32,
341
+ "retrieval_num_batches": 2,
342
+ "retrieval_num_samples": 64,
343
+ "retrieval_order_policy": "manifest_order",
344
+ "retrieval_unique_datasets_per_batch": {
345
+ "max": 1.0,
346
+ "mean": 1.0,
347
+ "min": 1.0,
348
+ "std": 0.0
349
+ }
350
+ }
351
+ }
352
+ },
353
+ "humanml_original_filtered_test": {
354
+ "embedding": {
355
+ "diversity_generated": 1.2619524338100758,
356
+ "motion_fid": 0.37949637116434376,
357
+ "motion_kid": {
358
+ "mean": 0.0001613574830927611,
359
+ "std": 0.0
360
+ },
361
+ "num_samples": 85
362
+ },
363
+ "num_samples": 85,
364
+ "physical": {
365
+ "body_jerk_mean": {
366
+ "max": 335.8555908203125,
367
+ "mean": 68.35952532151165,
368
+ "min": 16.2313232421875,
369
+ "std": 57.10007770878522
370
+ },
371
+ "contact_sliding_speed": {
372
+ "max": 1.2696173191070557,
373
+ "mean": 0.24481515460099806,
374
+ "min": 0.0,
375
+ "std": 0.26533379757226744
376
+ },
377
+ "foot_ground_error": {
378
+ "max": 0.22183531522750854,
379
+ "mean": 0.07665189722114626,
380
+ "min": 0.005901610944420099,
381
+ "std": 0.03276035910571276
382
+ },
383
+ "num_samples": 85
384
+ },
385
+ "text": {
386
+ "generated": {
387
+ "matching_distance": {
388
+ "max": 1.4400562047958374,
389
+ "mean": 1.119324480084812,
390
+ "min": 0.9264796376228333,
391
+ "std": 0.10692406599540519
392
+ },
393
+ "matching_score": 1.1193243265151978,
394
+ "mean_text_rank": 3.625,
395
+ "median_text_rank": 2.0,
396
+ "r_precision": [
397
+ 0.4375,
398
+ 0.59375,
399
+ 0.65625
400
+ ],
401
+ "retrieval_batch_size": 32,
402
+ "retrieval_num_batches": 2,
403
+ "retrieval_num_samples": 64,
404
+ "retrieval_order_policy": "manifest_order",
405
+ "retrieval_unique_datasets_per_batch": {
406
+ "max": 1.0,
407
+ "mean": 1.0,
408
+ "min": 1.0,
409
+ "std": 0.0
410
+ }
411
+ },
412
+ "reference": {
413
+ "matching_distance": {
414
+ "max": 1.3195350170135498,
415
+ "mean": 1.0584916128831752,
416
+ "min": 0.8765217065811157,
417
+ "std": 0.10127790732110797
418
+ },
419
+ "matching_score": 1.058491587638855,
420
+ "mean_text_rank": 2.71875,
421
+ "median_text_rank": 2.0,
422
+ "r_precision": [
423
+ 0.484375,
424
+ 0.671875,
425
+ 0.75
426
+ ],
427
+ "retrieval_batch_size": 32,
428
+ "retrieval_num_batches": 2,
429
+ "retrieval_num_samples": 64,
430
+ "retrieval_order_policy": "manifest_order",
431
+ "retrieval_unique_datasets_per_batch": {
432
+ "max": 1.0,
433
+ "mean": 1.0,
434
+ "min": 1.0,
435
+ "std": 0.0
436
+ }
437
+ }
438
+ }
439
+ },
440
+ "idea400_original_filtered_test": {
441
+ "embedding": {
442
+ "diversity_generated": 1.1603107092959224,
443
+ "motion_fid": 0.3089720542673662,
444
+ "motion_kid": {
445
+ "mean": 0.00024784405716316016,
446
+ "std": 0.0
447
+ },
448
+ "num_samples": 85
449
+ },
450
+ "num_samples": 85,
451
+ "physical": {
452
+ "body_jerk_mean": {
453
+ "max": 109.74834442138672,
454
+ "mean": 46.92937700608197,
455
+ "min": 12.396353721618652,
456
+ "std": 20.873435730428675
457
+ },
458
+ "contact_sliding_speed": {
459
+ "max": 0.8785831928253174,
460
+ "mean": 0.14769231055808418,
461
+ "min": 0.0,
462
+ "std": 0.24178951340825944
463
+ },
464
+ "foot_ground_error": {
465
+ "max": 0.2917259931564331,
466
+ "mean": 0.0424500749556019,
467
+ "min": 0.007580623496323824,
468
+ "std": 0.03828018100277618
469
+ },
470
+ "num_samples": 85
471
+ },
472
+ "text": {
473
+ "generated": {
474
+ "matching_distance": {
475
+ "max": 1.4350745677947998,
476
+ "mean": 1.1293431899126838,
477
+ "min": 0.9147061705589294,
478
+ "std": 0.10790011773020315
479
+ },
480
+ "matching_score": 1.1293431520462036,
481
+ "mean_text_rank": 7.203125,
482
+ "median_text_rank": 5.0,
483
+ "r_precision": [
484
+ 0.1875,
485
+ 0.34375,
486
+ 0.4375
487
+ ],
488
+ "retrieval_batch_size": 32,
489
+ "retrieval_num_batches": 2,
490
+ "retrieval_num_samples": 64,
491
+ "retrieval_order_policy": "manifest_order",
492
+ "retrieval_unique_datasets_per_batch": {
493
+ "max": 1.0,
494
+ "mean": 1.0,
495
+ "min": 1.0,
496
+ "std": 0.0
497
+ }
498
+ },
499
+ "reference": {
500
+ "matching_distance": {
501
+ "max": 1.3933959007263184,
502
+ "mean": 1.0893783358966602,
503
+ "min": 0.8957217931747437,
504
+ "std": 0.09054021801661577
505
+ },
506
+ "matching_score": 1.0893783569335938,
507
+ "mean_text_rank": 6.359375,
508
+ "median_text_rank": 4.0,
509
+ "r_precision": [
510
+ 0.21875,
511
+ 0.359375,
512
+ 0.46875
513
+ ],
514
+ "retrieval_batch_size": 32,
515
+ "retrieval_num_batches": 2,
516
+ "retrieval_num_samples": 64,
517
+ "retrieval_order_policy": "manifest_order",
518
+ "retrieval_unique_datasets_per_batch": {
519
+ "max": 1.0,
520
+ "mean": 1.0,
521
+ "min": 1.0,
522
+ "std": 0.0
523
+ }
524
+ }
525
+ }
526
+ },
527
+ "lafan1_original_filtered_test": {
528
+ "embedding": {
529
+ "diversity_generated": 1.0652331521765446,
530
+ "motion_fid": 0.4122437081306951,
531
+ "motion_kid": {
532
+ "mean": 0.00044010696667928784,
533
+ "std": 0.0
534
+ },
535
+ "num_samples": 85
536
+ },
537
+ "num_samples": 85,
538
+ "physical": {
539
+ "body_jerk_mean": {
540
+ "max": 314.1806945800781,
541
+ "mean": 139.9726458381204,
542
+ "min": 31.704696655273438,
543
+ "std": 71.70045395223674
544
+ },
545
+ "contact_sliding_speed": {
546
+ "max": 2.5723462104797363,
547
+ "mean": 1.074749160612769,
548
+ "min": 0.008806232362985611,
549
+ "std": 0.6605997776393984
550
+ },
551
+ "foot_ground_error": {
552
+ "max": 0.11140980571508408,
553
+ "mean": 0.021036932756781906,
554
+ "min": 0.00042374650365673006,
555
+ "std": 0.020322981130113965
556
+ },
557
+ "num_samples": 85
558
+ },
559
+ "text": {
560
+ "generated": {
561
+ "matching_distance": {
562
+ "max": 1.44863760471344,
563
+ "mean": 1.1349114824743831,
564
+ "min": 0.935748815536499,
565
+ "std": 0.10606785314295356
566
+ },
567
+ "matching_score": 1.1349115371704102,
568
+ "mean_text_rank": 8.84375,
569
+ "median_text_rank": 7.5,
570
+ "r_precision": [
571
+ 0.328125,
572
+ 0.359375,
573
+ 0.421875
574
+ ],
575
+ "retrieval_batch_size": 32,
576
+ "retrieval_num_batches": 2,
577
+ "retrieval_num_samples": 64,
578
+ "retrieval_order_policy": "manifest_order",
579
+ "retrieval_unique_datasets_per_batch": {
580
+ "max": 1.0,
581
+ "mean": 1.0,
582
+ "min": 1.0,
583
+ "std": 0.0
584
+ }
585
+ },
586
+ "reference": {
587
+ "matching_distance": {
588
+ "max": 1.4183552265167236,
589
+ "mean": 1.1475334244615891,
590
+ "min": 0.9194038510322571,
591
+ "std": 0.09896612738362433
592
+ },
593
+ "matching_score": 1.1475334167480469,
594
+ "mean_text_rank": 8.109375,
595
+ "median_text_rank": 5.0,
596
+ "r_precision": [
597
+ 0.359375,
598
+ 0.359375,
599
+ 0.375
600
+ ],
601
+ "retrieval_batch_size": 32,
602
+ "retrieval_num_batches": 2,
603
+ "retrieval_num_samples": 64,
604
+ "retrieval_order_policy": "manifest_order",
605
+ "retrieval_unique_datasets_per_batch": {
606
+ "max": 1.0,
607
+ "mean": 1.0,
608
+ "min": 1.0,
609
+ "std": 0.0
610
+ }
611
+ }
612
+ }
613
+ },
614
+ "motiongv_original_filtered_test": {
615
+ "embedding": {
616
+ "diversity_generated": 1.3252344922258321,
617
+ "motion_fid": 0.3607463810043918,
618
+ "motion_kid": {
619
+ "mean": 1.8889422511048082e-05,
620
+ "std": 0.0
621
+ },
622
+ "num_samples": 85
623
+ },
624
+ "num_samples": 85,
625
+ "physical": {
626
+ "body_jerk_mean": {
627
+ "max": 394.6324157714844,
628
+ "mean": 63.257873658572926,
629
+ "min": 10.021343231201172,
630
+ "std": 58.67747854537708
631
+ },
632
+ "contact_sliding_speed": {
633
+ "max": 0.7465254664421082,
634
+ "mean": 0.12326823623905725,
635
+ "min": 0.003070311388000846,
636
+ "std": 0.17676128678585645
637
+ },
638
+ "foot_ground_error": {
639
+ "max": 0.1117943823337555,
640
+ "mean": 0.07272094489239594,
641
+ "min": 0.012083700858056545,
642
+ "std": 0.01693819111508039
643
+ },
644
+ "num_samples": 85
645
+ },
646
+ "text": {
647
+ "generated": {
648
+ "matching_distance": {
649
+ "max": 1.4481498003005981,
650
+ "mean": 1.1457594983717974,
651
+ "min": 0.8835017085075378,
652
+ "std": 0.10666800116978434
653
+ },
654
+ "matching_score": 1.1457593441009521,
655
+ "mean_text_rank": 3.875,
656
+ "median_text_rank": 2.0,
657
+ "r_precision": [
658
+ 0.46875,
659
+ 0.59375,
660
+ 0.703125
661
+ ],
662
+ "retrieval_batch_size": 32,
663
+ "retrieval_num_batches": 2,
664
+ "retrieval_num_samples": 64,
665
+ "retrieval_order_policy": "manifest_order",
666
+ "retrieval_unique_datasets_per_batch": {
667
+ "max": 1.0,
668
+ "mean": 1.0,
669
+ "min": 1.0,
670
+ "std": 0.0
671
+ }
672
+ },
673
+ "reference": {
674
+ "matching_distance": {
675
+ "max": 1.3658052682876587,
676
+ "mean": 1.1116458219640395,
677
+ "min": 0.8830475211143494,
678
+ "std": 0.08860567172404131
679
+ },
680
+ "matching_score": 1.1116458177566528,
681
+ "mean_text_rank": 2.40625,
682
+ "median_text_rank": 1.0,
683
+ "r_precision": [
684
+ 0.609375,
685
+ 0.703125,
686
+ 0.828125
687
+ ],
688
+ "retrieval_batch_size": 32,
689
+ "retrieval_num_batches": 2,
690
+ "retrieval_num_samples": 64,
691
+ "retrieval_order_policy": "manifest_order",
692
+ "retrieval_unique_datasets_per_batch": {
693
+ "max": 1.0,
694
+ "mean": 1.0,
695
+ "min": 1.0,
696
+ "std": 0.0
697
+ }
698
+ }
699
+ }
700
+ },
701
+ "motionllama_original_filtered_test": {
702
+ "embedding": {
703
+ "diversity_generated": 1.243081346203865,
704
+ "motion_fid": 0.41212859468716423,
705
+ "motion_kid": {
706
+ "mean": 0.00039907121682780655,
707
+ "std": 0.0
708
+ },
709
+ "num_samples": 85
710
+ },
711
+ "num_samples": 85,
712
+ "physical": {
713
+ "body_jerk_mean": {
714
+ "max": 149.49559020996094,
715
+ "mean": 51.222771689471074,
716
+ "min": 11.60899543762207,
717
+ "std": 21.585048177328567
718
+ },
719
+ "contact_sliding_speed": {
720
+ "max": 0.3608182966709137,
721
+ "mean": 0.07486999163623242,
722
+ "min": 0.004085029475390911,
723
+ "std": 0.07851552799603706
724
+ },
725
+ "foot_ground_error": {
726
+ "max": 0.10084258019924164,
727
+ "mean": 0.06930571288308676,
728
+ "min": 0.012543153017759323,
729
+ "std": 0.017269924728060666
730
+ },
731
+ "num_samples": 85
732
+ },
733
+ "text": {
734
+ "generated": {
735
+ "matching_distance": {
736
+ "max": 1.3583039045333862,
737
+ "mean": 1.129518364457523,
738
+ "min": 0.907444417476654,
739
+ "std": 0.11405018318730081
740
+ },
741
+ "matching_score": 1.1295183897018433,
742
+ "mean_text_rank": 9.53125,
743
+ "median_text_rank": 7.0,
744
+ "r_precision": [
745
+ 0.171875,
746
+ 0.25,
747
+ 0.28125
748
+ ],
749
+ "retrieval_batch_size": 32,
750
+ "retrieval_num_batches": 2,
751
+ "retrieval_num_samples": 64,
752
+ "retrieval_order_policy": "manifest_order",
753
+ "retrieval_unique_datasets_per_batch": {
754
+ "max": 1.0,
755
+ "mean": 1.0,
756
+ "min": 1.0,
757
+ "std": 0.0
758
+ }
759
+ },
760
+ "reference": {
761
+ "matching_distance": {
762
+ "max": 1.2308001518249512,
763
+ "mean": 1.0292582687209635,
764
+ "min": 0.8811001777648926,
765
+ "std": 0.0877342679085498
766
+ },
767
+ "matching_score": 1.029258131980896,
768
+ "mean_text_rank": 6.921875,
769
+ "median_text_rank": 4.0,
770
+ "r_precision": [
771
+ 0.203125,
772
+ 0.3125,
773
+ 0.46875
774
+ ],
775
+ "retrieval_batch_size": 32,
776
+ "retrieval_num_batches": 2,
777
+ "retrieval_num_samples": 64,
778
+ "retrieval_order_policy": "manifest_order",
779
+ "retrieval_unique_datasets_per_batch": {
780
+ "max": 1.0,
781
+ "mean": 1.0,
782
+ "min": 1.0,
783
+ "std": 0.0
784
+ }
785
+ }
786
+ }
787
+ },
788
+ "omomo_original_filtered_test": {
789
+ "embedding": {
790
+ "diversity_generated": 1.225772022315077,
791
+ "motion_fid": 0.4057685475174302,
792
+ "motion_kid": {
793
+ "mean": 0.00022337189809062963,
794
+ "std": 0.0
795
+ },
796
+ "num_samples": 85
797
+ },
798
+ "num_samples": 85,
799
+ "physical": {
800
+ "body_jerk_mean": {
801
+ "max": 170.46762084960938,
802
+ "mean": 58.10617489534266,
803
+ "min": 16.497365951538086,
804
+ "std": 24.335758018848132
805
+ },
806
+ "contact_sliding_speed": {
807
+ "max": 1.117565631866455,
808
+ "mean": 0.3834191459504997,
809
+ "min": 0.005078582558780909,
810
+ "std": 0.3170930006320161
811
+ },
812
+ "foot_ground_error": {
813
+ "max": 0.1168566644191742,
814
+ "mean": 0.016199057431453293,
815
+ "min": 0.002339268336072564,
816
+ "std": 0.013331415234509801
817
+ },
818
+ "num_samples": 85
819
+ },
820
+ "text": {
821
+ "generated": {
822
+ "matching_distance": {
823
+ "max": 1.4282751083374023,
824
+ "mean": 1.1995593414587133,
825
+ "min": 0.970937192440033,
826
+ "std": 0.09437652461416668
827
+ },
828
+ "matching_score": 1.1995593309402466,
829
+ "mean_text_rank": 9.390625,
830
+ "median_text_rank": 7.0,
831
+ "r_precision": [
832
+ 0.234375,
833
+ 0.328125,
834
+ 0.34375
835
+ ],
836
+ "retrieval_batch_size": 32,
837
+ "retrieval_num_batches": 2,
838
+ "retrieval_num_samples": 64,
839
+ "retrieval_order_policy": "manifest_order",
840
+ "retrieval_unique_datasets_per_batch": {
841
+ "max": 1.0,
842
+ "mean": 1.0,
843
+ "min": 1.0,
844
+ "std": 0.0
845
+ }
846
+ },
847
+ "reference": {
848
+ "matching_distance": {
849
+ "max": 1.3254883289337158,
850
+ "mean": 1.1843213705455555,
851
+ "min": 0.958416759967804,
852
+ "std": 0.07360095182352315
853
+ },
854
+ "matching_score": 1.1843212842941284,
855
+ "mean_text_rank": 7.484375,
856
+ "median_text_rank": 5.5,
857
+ "r_precision": [
858
+ 0.140625,
859
+ 0.328125,
860
+ 0.40625
861
+ ],
862
+ "retrieval_batch_size": 32,
863
+ "retrieval_num_batches": 2,
864
+ "retrieval_num_samples": 64,
865
+ "retrieval_order_policy": "manifest_order",
866
+ "retrieval_unique_datasets_per_batch": {
867
+ "max": 1.0,
868
+ "mean": 1.0,
869
+ "min": 1.0,
870
+ "std": 0.0
871
+ }
872
+ }
873
+ }
874
+ },
875
+ "permo_original_filtered_test": {
876
+ "embedding": {
877
+ "diversity_generated": 1.2086325566904998,
878
+ "motion_fid": 0.4148669497066708,
879
+ "motion_kid": {
880
+ "mean": 0.0003991512469281311,
881
+ "std": 0.0
882
+ },
883
+ "num_samples": 85
884
+ },
885
+ "num_samples": 85,
886
+ "physical": {
887
+ "body_jerk_mean": {
888
+ "max": 240.15951538085938,
889
+ "mean": 89.35240687202005,
890
+ "min": 24.114988327026367,
891
+ "std": 42.25908942517282
892
+ },
893
+ "contact_sliding_speed": {
894
+ "max": 1.3213852643966675,
895
+ "mean": 0.38401777897687517,
896
+ "min": 0.0,
897
+ "std": 0.3190618849745317
898
+ },
899
+ "foot_ground_error": {
900
+ "max": 0.2258841097354889,
901
+ "mean": 0.036138936230803234,
902
+ "min": 0.008152862079441547,
903
+ "std": 0.02827083711142457
904
+ },
905
+ "num_samples": 85
906
+ },
907
+ "text": {
908
+ "generated": {
909
+ "matching_distance": {
910
+ "max": 1.412964940071106,
911
+ "mean": 1.104096428085776,
912
+ "min": 0.8289653062820435,
913
+ "std": 0.12450774868152055
914
+ },
915
+ "matching_score": 1.1040964126586914,
916
+ "mean_text_rank": 5.796875,
917
+ "median_text_rank": 4.0,
918
+ "r_precision": [
919
+ 0.15625,
920
+ 0.359375,
921
+ 0.4375
922
+ ],
923
+ "retrieval_batch_size": 32,
924
+ "retrieval_num_batches": 2,
925
+ "retrieval_num_samples": 64,
926
+ "retrieval_order_policy": "manifest_order",
927
+ "retrieval_unique_datasets_per_batch": {
928
+ "max": 1.0,
929
+ "mean": 1.0,
930
+ "min": 1.0,
931
+ "std": 0.0
932
+ }
933
+ },
934
+ "reference": {
935
+ "matching_distance": {
936
+ "max": 1.2057435512542725,
937
+ "mean": 1.0011776215889874,
938
+ "min": 0.8405618071556091,
939
+ "std": 0.0876302018200968
940
+ },
941
+ "matching_score": 1.0011776685714722,
942
+ "mean_text_rank": 4.84375,
943
+ "median_text_rank": 4.0,
944
+ "r_precision": [
945
+ 0.171875,
946
+ 0.328125,
947
+ 0.46875
948
+ ],
949
+ "retrieval_batch_size": 32,
950
+ "retrieval_num_batches": 2,
951
+ "retrieval_num_samples": 64,
952
+ "retrieval_order_policy": "manifest_order",
953
+ "retrieval_unique_datasets_per_batch": {
954
+ "max": 1.0,
955
+ "mean": 1.0,
956
+ "min": 1.0,
957
+ "std": 0.0
958
+ }
959
+ }
960
+ }
961
+ },
962
+ "snapmogen_original_filtered_test": {
963
+ "embedding": {
964
+ "diversity_generated": 1.1717517212600463,
965
+ "motion_fid": 0.3023945940485344,
966
+ "motion_kid": {
967
+ "mean": 0.0002671196633574624,
968
+ "std": 0.0
969
+ },
970
+ "num_samples": 85
971
+ },
972
+ "num_samples": 85,
973
+ "physical": {
974
+ "body_jerk_mean": {
975
+ "max": 183.62680053710938,
976
+ "mean": 57.84897658404182,
977
+ "min": 16.209308624267578,
978
+ "std": 38.09984754580235
979
+ },
980
+ "contact_sliding_speed": {
981
+ "max": 1.0688523054122925,
982
+ "mean": 0.1868419358452015,
983
+ "min": 0.0041826097294688225,
984
+ "std": 0.20431294637395492
985
+ },
986
+ "foot_ground_error": {
987
+ "max": 0.14229609072208405,
988
+ "mean": 0.034169711305431144,
989
+ "min": 0.0016218408709391952,
990
+ "std": 0.02866694918161823
991
+ },
992
+ "num_samples": 85
993
+ },
994
+ "text": {
995
+ "generated": {
996
+ "matching_distance": {
997
+ "max": 1.3117051124572754,
998
+ "mean": 1.0969315627041984,
999
+ "min": 0.9296107888221741,
1000
+ "std": 0.08918784110822944
1001
+ },
1002
+ "matching_score": 1.0969316959381104,
1003
+ "mean_text_rank": 4.265625,
1004
+ "median_text_rank": 3.0,
1005
+ "r_precision": [
1006
+ 0.359375,
1007
+ 0.46875,
1008
+ 0.65625
1009
+ ],
1010
+ "retrieval_batch_size": 32,
1011
+ "retrieval_num_batches": 2,
1012
+ "retrieval_num_samples": 64,
1013
+ "retrieval_order_policy": "manifest_order",
1014
+ "retrieval_unique_datasets_per_batch": {
1015
+ "max": 1.0,
1016
+ "mean": 1.0,
1017
+ "min": 1.0,
1018
+ "std": 0.0
1019
+ }
1020
+ },
1021
+ "reference": {
1022
+ "matching_distance": {
1023
+ "max": 1.2125790119171143,
1024
+ "mean": 1.0510956680073458,
1025
+ "min": 0.8823143839836121,
1026
+ "std": 0.08172085751280951
1027
+ },
1028
+ "matching_score": 1.051095724105835,
1029
+ "mean_text_rank": 3.140625,
1030
+ "median_text_rank": 2.0,
1031
+ "r_precision": [
1032
+ 0.46875,
1033
+ 0.671875,
1034
+ 0.765625
1035
+ ],
1036
+ "retrieval_batch_size": 32,
1037
+ "retrieval_num_batches": 2,
1038
+ "retrieval_num_samples": 64,
1039
+ "retrieval_order_policy": "manifest_order",
1040
+ "retrieval_unique_datasets_per_batch": {
1041
+ "max": 1.0,
1042
+ "mean": 1.0,
1043
+ "min": 1.0,
1044
+ "std": 0.0
1045
+ }
1046
+ }
1047
+ }
1048
+ }
1049
+ },
1050
+ "embedding_frames": 120,
1051
+ "embedding_metrics": {
1052
+ "diversity_generated": 1.3800554760633976,
1053
+ "embedding_frames": 120,
1054
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
1055
+ "matching_score_generated": 1.1404733657836914,
1056
+ "matching_score_reference": 1.0944490432739258,
1057
+ "motion_fid": 0.10967632126673776,
1058
+ "motion_key": "body_pos_local",
1059
+ "motion_kid": {
1060
+ "mean": 7.427152145522697e-05,
1061
+ "std": 0.0
1062
+ },
1063
+ "multimodality": null,
1064
+ "num_samples": 1024,
1065
+ "r_precision_generated": [
1066
+ 0.5869140625,
1067
+ 0.7333984375,
1068
+ 0.818359375
1069
+ ],
1070
+ "r_precision_reference": [
1071
+ 0.6611328125,
1072
+ 0.82421875,
1073
+ 0.8916015625
1074
+ ],
1075
+ "text_cache_dir": "${OMG_REPO_ROOT}/outputs/evaluator_text_cache/filtered_original_g1_tmr_bodypos_60f",
1076
+ "text_encoder_model": "${OMG_REPO_ROOT}/models/t5-3b-local",
1077
+ "text_generated": {
1078
+ "matching_distance": {
1079
+ "max": 1.5543382167816162,
1080
+ "mean": 1.1404733364470303,
1081
+ "min": 0.8289653062820435,
1082
+ "std": 0.10747345553020096
1083
+ },
1084
+ "matching_score": 1.1404733657836914,
1085
+ "mean_text_rank": 2.576171875,
1086
+ "median_text_rank": 1.0,
1087
+ "r_precision": [
1088
+ 0.5869140625,
1089
+ 0.7333984375,
1090
+ 0.818359375
1091
+ ],
1092
+ "retrieval_batch_size": 32,
1093
+ "retrieval_num_batches": 32,
1094
+ "retrieval_num_samples": 1024,
1095
+ "retrieval_order_policy": "stratified_by_dataset",
1096
+ "retrieval_unique_datasets_per_batch": {
1097
+ "max": 12.0,
1098
+ "mean": 11.9375,
1099
+ "min": 10.0,
1100
+ "std": 0.34798527267687634
1101
+ }
1102
+ },
1103
+ "text_reference": {
1104
+ "matching_distance": {
1105
+ "max": 1.4848475456237793,
1106
+ "mean": 1.0944489734247327,
1107
+ "min": 0.8405618071556091,
1108
+ "std": 0.10523510496847523
1109
+ },
1110
+ "matching_score": 1.0944490432739258,
1111
+ "mean_text_rank": 1.888671875,
1112
+ "median_text_rank": 1.0,
1113
+ "r_precision": [
1114
+ 0.6611328125,
1115
+ 0.82421875,
1116
+ 0.8916015625
1117
+ ],
1118
+ "retrieval_batch_size": 32,
1119
+ "retrieval_num_batches": 32,
1120
+ "retrieval_num_samples": 1024,
1121
+ "retrieval_order_policy": "stratified_by_dataset",
1122
+ "retrieval_unique_datasets_per_batch": {
1123
+ "max": 12.0,
1124
+ "mean": 11.9375,
1125
+ "min": 10.0,
1126
+ "std": 0.34798527267687634
1127
+ }
1128
+ },
1129
+ "text_retrieval_batch_size": 32
1130
+ },
1131
+ "exp": "300m",
1132
+ "generated_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/generated_qpos_raw.npz",
1133
+ "label": "OmniMotion 300M mixed-modal scaling s050000",
1134
+ "num_frames": 120,
1135
+ "num_samples": 1024,
1136
+ "physical_metrics": {
1137
+ "body_jerk_mean": {
1138
+ "max": 394.6324157714844,
1139
+ "mean": 74.01262681838125,
1140
+ "min": 10.021343231201172,
1141
+ "std": 53.226579084617505
1142
+ },
1143
+ "contact_height_threshold": 0.12,
1144
+ "contact_penetration_tolerance": 0.02,
1145
+ "contact_sliding_speed": {
1146
+ "max": 2.5723462104797363,
1147
+ "mean": 0.35662794705444867,
1148
+ "min": 0.0,
1149
+ "std": 0.4447682159597098
1150
+ },
1151
+ "foot_ground_error": {
1152
+ "max": 0.2917259931564331,
1153
+ "mean": 0.04151647091404698,
1154
+ "min": 0.00042374650365673006,
1155
+ "std": 0.03340117411035358
1156
+ },
1157
+ "frames": 120,
1158
+ "num_samples": 1024
1159
+ },
1160
+ "real_motion_metrics": {
1161
+ "diversity_real": 1.3768246488138,
1162
+ "matching_distance_real": {
1163
+ "max": 1.4848475456237793,
1164
+ "mean": 1.0944489734247327,
1165
+ "min": 0.8405618071556091,
1166
+ "std": 0.10523510496847523
1167
+ },
1168
+ "matching_score_real": 1.0944490432739258,
1169
+ "motion_fid_reference_eval": 0.0,
1170
+ "motion_fid_split_half": 0.0,
1171
+ "motion_kid_reference_eval": {
1172
+ "mean": -1.0749719189018236e-05,
1173
+ "std": 0.0
1174
+ },
1175
+ "motion_kid_split_half": {
1176
+ "mean": -1.0749719189018236e-05,
1177
+ "std": 0.0
1178
+ },
1179
+ "multimodality": "-",
1180
+ "num_reference_samples": 1024,
1181
+ "num_samples": 1024,
1182
+ "r_precision_real": [
1183
+ 0.3740234375,
1184
+ 0.5009765625,
1185
+ 0.5810546875
1186
+ ],
1187
+ "text_real": {
1188
+ "matching_distance": {
1189
+ "max": 1.4848475456237793,
1190
+ "mean": 1.0944489734247327,
1191
+ "min": 0.8405618071556091,
1192
+ "std": 0.10523510496847523
1193
+ },
1194
+ "matching_score": 1.0944490432739258,
1195
+ "mean_text_rank": 5.232421875,
1196
+ "median_text_rank": 2.0,
1197
+ "r_precision": [
1198
+ 0.3740234375,
1199
+ 0.5009765625,
1200
+ 0.5810546875
1201
+ ],
1202
+ "retrieval_batch_size": 32,
1203
+ "retrieval_num_batches": 32,
1204
+ "retrieval_num_samples": 1024,
1205
+ "retrieval_order_policy": "manifest_order",
1206
+ "retrieval_unique_datasets_per_batch": null
1207
+ }
1208
+ },
1209
+ "reference_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/reference_qpos.npz",
1210
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/sample_metrics.jsonl",
1211
+ "sample_rankings_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/sample_rankings.json",
1212
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_samples/mixed_modalities_all_v1/text_test_1024.jsonl",
1213
+ "split": "test",
1214
+ "tracker_executed": {
1215
+ "action_clip": 10.0,
1216
+ "artifact_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/tracker_executed/tracker_executed_qpos.npz",
1217
+ "benchmark_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/tracker_executed/benchmark_tracker.json",
1218
+ "control_substeps": 10,
1219
+ "enabled": true,
1220
+ "holomotion_onnx": "${OMG_MODELS_ROOT}/HoloMotion/HoloMotion_v1.2/holomotion_v1.2_motion_tracking_model/exported/model.onnx",
1221
+ "metrics": {
1222
+ "e_acc": {
1223
+ "max": 5.753578186035156,
1224
+ "mean": 1.3608852332108654,
1225
+ "min": 0.19200170040130615,
1226
+ "std": 1.0193065179674143
1227
+ },
1228
+ "e_vel": {
1229
+ "max": 27.857595443725586,
1230
+ "mean": 4.456644410674926,
1231
+ "min": 0.6209603548049927,
1232
+ "std": 3.3743479287514506
1233
+ },
1234
+ "enabled": true,
1235
+ "g_mpjpe": {
1236
+ "max": 1602.4024658203125,
1237
+ "mean": 130.28732685372233,
1238
+ "min": 18.030338287353516,
1239
+ "std": 129.4890607920233
1240
+ },
1241
+ "mpjpe": {
1242
+ "max": 316.797119140625,
1243
+ "mean": 45.63376460596919,
1244
+ "min": 13.653532981872559,
1245
+ "std": 27.06205114115315
1246
+ },
1247
+ "num_samples": 1024,
1248
+ "root_index": 0,
1249
+ "target_fps": 50.0
1250
+ },
1251
+ "metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/tracker_executed/tracker_metrics.json",
1252
+ "num_generated_samples": 1024,
1253
+ "num_samples": 1024,
1254
+ "providers": "TensorrtExecutionProvider,CUDAExecutionProvider",
1255
+ "robot_xml": null,
1256
+ "root_index": 0,
1257
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/tracker_executed/tracker_sample_metrics.jsonl",
1258
+ "selected_indices": [
1259
+ 0,
1260
+ 1,
1261
+ 2,
1262
+ 3,
1263
+ 4,
1264
+ 5,
1265
+ 6,
1266
+ 7,
1267
+ 8,
1268
+ 9,
1269
+ 10,
1270
+ 11,
1271
+ 12,
1272
+ 13,
1273
+ 14,
1274
+ 15,
1275
+ 16,
1276
+ 17,
1277
+ 18,
1278
+ 19,
1279
+ 20,
1280
+ 21,
1281
+ 22,
1282
+ 23,
1283
+ 24,
1284
+ 25,
1285
+ 26,
1286
+ 27,
1287
+ 28,
1288
+ 29,
1289
+ 30,
1290
+ 31,
1291
+ 32,
1292
+ 33,
1293
+ 34,
1294
+ 35,
1295
+ 36,
1296
+ 37,
1297
+ 38,
1298
+ 39,
1299
+ 40,
1300
+ 41,
1301
+ 42,
1302
+ 43,
1303
+ 44,
1304
+ 45,
1305
+ 46,
1306
+ 47,
1307
+ 48,
1308
+ 49,
1309
+ 50,
1310
+ 51,
1311
+ 52,
1312
+ 53,
1313
+ 54,
1314
+ 55,
1315
+ 56,
1316
+ 57,
1317
+ 58,
1318
+ 59,
1319
+ 60,
1320
+ 61,
1321
+ 62,
1322
+ 63,
1323
+ 64,
1324
+ 65,
1325
+ 66,
1326
+ 67,
1327
+ 68,
1328
+ 69,
1329
+ 70,
1330
+ 71,
1331
+ 72,
1332
+ 73,
1333
+ 74,
1334
+ 75,
1335
+ 76,
1336
+ 77,
1337
+ 78,
1338
+ 79,
1339
+ 80,
1340
+ 81,
1341
+ 82,
1342
+ 83,
1343
+ 84,
1344
+ 85,
1345
+ 86,
1346
+ 87,
1347
+ 88,
1348
+ 89,
1349
+ 90,
1350
+ 91,
1351
+ 92,
1352
+ 93,
1353
+ 94,
1354
+ 95,
1355
+ 96,
1356
+ 97,
1357
+ 98,
1358
+ 99,
1359
+ 100,
1360
+ 101,
1361
+ 102,
1362
+ 103,
1363
+ 104,
1364
+ 105,
1365
+ 106,
1366
+ 107,
1367
+ 108,
1368
+ 109,
1369
+ 110,
1370
+ 111,
1371
+ 112,
1372
+ 113,
1373
+ 114,
1374
+ 115,
1375
+ 116,
1376
+ 117,
1377
+ 118,
1378
+ 119,
1379
+ 120,
1380
+ 121,
1381
+ 122,
1382
+ 123,
1383
+ 124,
1384
+ 125,
1385
+ 126,
1386
+ 127,
1387
+ 128,
1388
+ 129,
1389
+ 130,
1390
+ 131,
1391
+ 132,
1392
+ 133,
1393
+ 134,
1394
+ 135,
1395
+ 136,
1396
+ 137,
1397
+ 138,
1398
+ 139,
1399
+ 140,
1400
+ 141,
1401
+ 142,
1402
+ 143,
1403
+ 144,
1404
+ 145,
1405
+ 146,
1406
+ 147,
1407
+ 148,
1408
+ 149,
1409
+ 150,
1410
+ 151,
1411
+ 152,
1412
+ 153,
1413
+ 154,
1414
+ 155,
1415
+ 156,
1416
+ 157,
1417
+ 158,
1418
+ 159,
1419
+ 160,
1420
+ 161,
1421
+ 162,
1422
+ 163,
1423
+ 164,
1424
+ 165,
1425
+ 166,
1426
+ 167,
1427
+ 168,
1428
+ 169,
1429
+ 170,
1430
+ 171,
1431
+ 172,
1432
+ 173,
1433
+ 174,
1434
+ 175,
1435
+ 176,
1436
+ 177,
1437
+ 178,
1438
+ 179,
1439
+ 180,
1440
+ 181,
1441
+ 182,
1442
+ 183,
1443
+ 184,
1444
+ 185,
1445
+ 186,
1446
+ 187,
1447
+ 188,
1448
+ 189,
1449
+ 190,
1450
+ 191,
1451
+ 192,
1452
+ 193,
1453
+ 194,
1454
+ 195,
1455
+ 196,
1456
+ 197,
1457
+ 198,
1458
+ 199,
1459
+ 200,
1460
+ 201,
1461
+ 202,
1462
+ 203,
1463
+ 204,
1464
+ 205,
1465
+ 206,
1466
+ 207,
1467
+ 208,
1468
+ 209,
1469
+ 210,
1470
+ 211,
1471
+ 212,
1472
+ 213,
1473
+ 214,
1474
+ 215,
1475
+ 216,
1476
+ 217,
1477
+ 218,
1478
+ 219,
1479
+ 220,
1480
+ 221,
1481
+ 222,
1482
+ 223,
1483
+ 224,
1484
+ 225,
1485
+ 226,
1486
+ 227,
1487
+ 228,
1488
+ 229,
1489
+ 230,
1490
+ 231,
1491
+ 232,
1492
+ 233,
1493
+ 234,
1494
+ 235,
1495
+ 236,
1496
+ 237,
1497
+ 238,
1498
+ 239,
1499
+ 240,
1500
+ 241,
1501
+ 242,
1502
+ 243,
1503
+ 244,
1504
+ 245,
1505
+ 246,
1506
+ 247,
1507
+ 248,
1508
+ 249,
1509
+ 250,
1510
+ 251,
1511
+ 252,
1512
+ 253,
1513
+ 254,
1514
+ 255,
1515
+ 256,
1516
+ 257,
1517
+ 258,
1518
+ 259,
1519
+ 260,
1520
+ 261,
1521
+ 262,
1522
+ 263,
1523
+ 264,
1524
+ 265,
1525
+ 266,
1526
+ 267,
1527
+ 268,
1528
+ 269,
1529
+ 270,
1530
+ 271,
1531
+ 272,
1532
+ 273,
1533
+ 274,
1534
+ 275,
1535
+ 276,
1536
+ 277,
1537
+ 278,
1538
+ 279,
1539
+ 280,
1540
+ 281,
1541
+ 282,
1542
+ 283,
1543
+ 284,
1544
+ 285,
1545
+ 286,
1546
+ 287,
1547
+ 288,
1548
+ 289,
1549
+ 290,
1550
+ 291,
1551
+ 292,
1552
+ 293,
1553
+ 294,
1554
+ 295,
1555
+ 296,
1556
+ 297,
1557
+ 298,
1558
+ 299,
1559
+ 300,
1560
+ 301,
1561
+ 302,
1562
+ 303,
1563
+ 304,
1564
+ 305,
1565
+ 306,
1566
+ 307,
1567
+ 308,
1568
+ 309,
1569
+ 310,
1570
+ 311,
1571
+ 312,
1572
+ 313,
1573
+ 314,
1574
+ 315,
1575
+ 316,
1576
+ 317,
1577
+ 318,
1578
+ 319,
1579
+ 320,
1580
+ 321,
1581
+ 322,
1582
+ 323,
1583
+ 324,
1584
+ 325,
1585
+ 326,
1586
+ 327,
1587
+ 328,
1588
+ 329,
1589
+ 330,
1590
+ 331,
1591
+ 332,
1592
+ 333,
1593
+ 334,
1594
+ 335,
1595
+ 336,
1596
+ 337,
1597
+ 338,
1598
+ 339,
1599
+ 340,
1600
+ 341,
1601
+ 342,
1602
+ 343,
1603
+ 344,
1604
+ 345,
1605
+ 346,
1606
+ 347,
1607
+ 348,
1608
+ 349,
1609
+ 350,
1610
+ 351,
1611
+ 352,
1612
+ 353,
1613
+ 354,
1614
+ 355,
1615
+ 356,
1616
+ 357,
1617
+ 358,
1618
+ 359,
1619
+ 360,
1620
+ 361,
1621
+ 362,
1622
+ 363,
1623
+ 364,
1624
+ 365,
1625
+ 366,
1626
+ 367,
1627
+ 368,
1628
+ 369,
1629
+ 370,
1630
+ 371,
1631
+ 372,
1632
+ 373,
1633
+ 374,
1634
+ 375,
1635
+ 376,
1636
+ 377,
1637
+ 378,
1638
+ 379,
1639
+ 380,
1640
+ 381,
1641
+ 382,
1642
+ 383,
1643
+ 384,
1644
+ 385,
1645
+ 386,
1646
+ 387,
1647
+ 388,
1648
+ 389,
1649
+ 390,
1650
+ 391,
1651
+ 392,
1652
+ 393,
1653
+ 394,
1654
+ 395,
1655
+ 396,
1656
+ 397,
1657
+ 398,
1658
+ 399,
1659
+ 400,
1660
+ 401,
1661
+ 402,
1662
+ 403,
1663
+ 404,
1664
+ 405,
1665
+ 406,
1666
+ 407,
1667
+ 408,
1668
+ 409,
1669
+ 410,
1670
+ 411,
1671
+ 412,
1672
+ 413,
1673
+ 414,
1674
+ 415,
1675
+ 416,
1676
+ 417,
1677
+ 418,
1678
+ 419,
1679
+ 420,
1680
+ 421,
1681
+ 422,
1682
+ 423,
1683
+ 424,
1684
+ 425,
1685
+ 426,
1686
+ 427,
1687
+ 428,
1688
+ 429,
1689
+ 430,
1690
+ 431,
1691
+ 432,
1692
+ 433,
1693
+ 434,
1694
+ 435,
1695
+ 436,
1696
+ 437,
1697
+ 438,
1698
+ 439,
1699
+ 440,
1700
+ 441,
1701
+ 442,
1702
+ 443,
1703
+ 444,
1704
+ 445,
1705
+ 446,
1706
+ 447,
1707
+ 448,
1708
+ 449,
1709
+ 450,
1710
+ 451,
1711
+ 452,
1712
+ 453,
1713
+ 454,
1714
+ 455,
1715
+ 456,
1716
+ 457,
1717
+ 458,
1718
+ 459,
1719
+ 460,
1720
+ 461,
1721
+ 462,
1722
+ 463,
1723
+ 464,
1724
+ 465,
1725
+ 466,
1726
+ 467,
1727
+ 468,
1728
+ 469,
1729
+ 470,
1730
+ 471,
1731
+ 472,
1732
+ 473,
1733
+ 474,
1734
+ 475,
1735
+ 476,
1736
+ 477,
1737
+ 478,
1738
+ 479,
1739
+ 480,
1740
+ 481,
1741
+ 482,
1742
+ 483,
1743
+ 484,
1744
+ 485,
1745
+ 486,
1746
+ 487,
1747
+ 488,
1748
+ 489,
1749
+ 490,
1750
+ 491,
1751
+ 492,
1752
+ 493,
1753
+ 494,
1754
+ 495,
1755
+ 496,
1756
+ 497,
1757
+ 498,
1758
+ 499,
1759
+ 500,
1760
+ 501,
1761
+ 502,
1762
+ 503,
1763
+ 504,
1764
+ 505,
1765
+ 506,
1766
+ 507,
1767
+ 508,
1768
+ 509,
1769
+ 510,
1770
+ 511,
1771
+ 512,
1772
+ 513,
1773
+ 514,
1774
+ 515,
1775
+ 516,
1776
+ 517,
1777
+ 518,
1778
+ 519,
1779
+ 520,
1780
+ 521,
1781
+ 522,
1782
+ 523,
1783
+ 524,
1784
+ 525,
1785
+ 526,
1786
+ 527,
1787
+ 528,
1788
+ 529,
1789
+ 530,
1790
+ 531,
1791
+ 532,
1792
+ 533,
1793
+ 534,
1794
+ 535,
1795
+ 536,
1796
+ 537,
1797
+ 538,
1798
+ 539,
1799
+ 540,
1800
+ 541,
1801
+ 542,
1802
+ 543,
1803
+ 544,
1804
+ 545,
1805
+ 546,
1806
+ 547,
1807
+ 548,
1808
+ 549,
1809
+ 550,
1810
+ 551,
1811
+ 552,
1812
+ 553,
1813
+ 554,
1814
+ 555,
1815
+ 556,
1816
+ 557,
1817
+ 558,
1818
+ 559,
1819
+ 560,
1820
+ 561,
1821
+ 562,
1822
+ 563,
1823
+ 564,
1824
+ 565,
1825
+ 566,
1826
+ 567,
1827
+ 568,
1828
+ 569,
1829
+ 570,
1830
+ 571,
1831
+ 572,
1832
+ 573,
1833
+ 574,
1834
+ 575,
1835
+ 576,
1836
+ 577,
1837
+ 578,
1838
+ 579,
1839
+ 580,
1840
+ 581,
1841
+ 582,
1842
+ 583,
1843
+ 584,
1844
+ 585,
1845
+ 586,
1846
+ 587,
1847
+ 588,
1848
+ 589,
1849
+ 590,
1850
+ 591,
1851
+ 592,
1852
+ 593,
1853
+ 594,
1854
+ 595,
1855
+ 596,
1856
+ 597,
1857
+ 598,
1858
+ 599,
1859
+ 600,
1860
+ 601,
1861
+ 602,
1862
+ 603,
1863
+ 604,
1864
+ 605,
1865
+ 606,
1866
+ 607,
1867
+ 608,
1868
+ 609,
1869
+ 610,
1870
+ 611,
1871
+ 612,
1872
+ 613,
1873
+ 614,
1874
+ 615,
1875
+ 616,
1876
+ 617,
1877
+ 618,
1878
+ 619,
1879
+ 620,
1880
+ 621,
1881
+ 622,
1882
+ 623,
1883
+ 624,
1884
+ 625,
1885
+ 626,
1886
+ 627,
1887
+ 628,
1888
+ 629,
1889
+ 630,
1890
+ 631,
1891
+ 632,
1892
+ 633,
1893
+ 634,
1894
+ 635,
1895
+ 636,
1896
+ 637,
1897
+ 638,
1898
+ 639,
1899
+ 640,
1900
+ 641,
1901
+ 642,
1902
+ 643,
1903
+ 644,
1904
+ 645,
1905
+ 646,
1906
+ 647,
1907
+ 648,
1908
+ 649,
1909
+ 650,
1910
+ 651,
1911
+ 652,
1912
+ 653,
1913
+ 654,
1914
+ 655,
1915
+ 656,
1916
+ 657,
1917
+ 658,
1918
+ 659,
1919
+ 660,
1920
+ 661,
1921
+ 662,
1922
+ 663,
1923
+ 664,
1924
+ 665,
1925
+ 666,
1926
+ 667,
1927
+ 668,
1928
+ 669,
1929
+ 670,
1930
+ 671,
1931
+ 672,
1932
+ 673,
1933
+ 674,
1934
+ 675,
1935
+ 676,
1936
+ 677,
1937
+ 678,
1938
+ 679,
1939
+ 680,
1940
+ 681,
1941
+ 682,
1942
+ 683,
1943
+ 684,
1944
+ 685,
1945
+ 686,
1946
+ 687,
1947
+ 688,
1948
+ 689,
1949
+ 690,
1950
+ 691,
1951
+ 692,
1952
+ 693,
1953
+ 694,
1954
+ 695,
1955
+ 696,
1956
+ 697,
1957
+ 698,
1958
+ 699,
1959
+ 700,
1960
+ 701,
1961
+ 702,
1962
+ 703,
1963
+ 704,
1964
+ 705,
1965
+ 706,
1966
+ 707,
1967
+ 708,
1968
+ 709,
1969
+ 710,
1970
+ 711,
1971
+ 712,
1972
+ 713,
1973
+ 714,
1974
+ 715,
1975
+ 716,
1976
+ 717,
1977
+ 718,
1978
+ 719,
1979
+ 720,
1980
+ 721,
1981
+ 722,
1982
+ 723,
1983
+ 724,
1984
+ 725,
1985
+ 726,
1986
+ 727,
1987
+ 728,
1988
+ 729,
1989
+ 730,
1990
+ 731,
1991
+ 732,
1992
+ 733,
1993
+ 734,
1994
+ 735,
1995
+ 736,
1996
+ 737,
1997
+ 738,
1998
+ 739,
1999
+ 740,
2000
+ 741,
2001
+ 742,
2002
+ 743,
2003
+ 744,
2004
+ 745,
2005
+ 746,
2006
+ 747,
2007
+ 748,
2008
+ 749,
2009
+ 750,
2010
+ 751,
2011
+ 752,
2012
+ 753,
2013
+ 754,
2014
+ 755,
2015
+ 756,
2016
+ 757,
2017
+ 758,
2018
+ 759,
2019
+ 760,
2020
+ 761,
2021
+ 762,
2022
+ 763,
2023
+ 764,
2024
+ 765,
2025
+ 766,
2026
+ 767,
2027
+ 768,
2028
+ 769,
2029
+ 770,
2030
+ 771,
2031
+ 772,
2032
+ 773,
2033
+ 774,
2034
+ 775,
2035
+ 776,
2036
+ 777,
2037
+ 778,
2038
+ 779,
2039
+ 780,
2040
+ 781,
2041
+ 782,
2042
+ 783,
2043
+ 784,
2044
+ 785,
2045
+ 786,
2046
+ 787,
2047
+ 788,
2048
+ 789,
2049
+ 790,
2050
+ 791,
2051
+ 792,
2052
+ 793,
2053
+ 794,
2054
+ 795,
2055
+ 796,
2056
+ 797,
2057
+ 798,
2058
+ 799,
2059
+ 800,
2060
+ 801,
2061
+ 802,
2062
+ 803,
2063
+ 804,
2064
+ 805,
2065
+ 806,
2066
+ 807,
2067
+ 808,
2068
+ 809,
2069
+ 810,
2070
+ 811,
2071
+ 812,
2072
+ 813,
2073
+ 814,
2074
+ 815,
2075
+ 816,
2076
+ 817,
2077
+ 818,
2078
+ 819,
2079
+ 820,
2080
+ 821,
2081
+ 822,
2082
+ 823,
2083
+ 824,
2084
+ 825,
2085
+ 826,
2086
+ 827,
2087
+ 828,
2088
+ 829,
2089
+ 830,
2090
+ 831,
2091
+ 832,
2092
+ 833,
2093
+ 834,
2094
+ 835,
2095
+ 836,
2096
+ 837,
2097
+ 838,
2098
+ 839,
2099
+ 840,
2100
+ 841,
2101
+ 842,
2102
+ 843,
2103
+ 844,
2104
+ 845,
2105
+ 846,
2106
+ 847,
2107
+ 848,
2108
+ 849,
2109
+ 850,
2110
+ 851,
2111
+ 852,
2112
+ 853,
2113
+ 854,
2114
+ 855,
2115
+ 856,
2116
+ 857,
2117
+ 858,
2118
+ 859,
2119
+ 860,
2120
+ 861,
2121
+ 862,
2122
+ 863,
2123
+ 864,
2124
+ 865,
2125
+ 866,
2126
+ 867,
2127
+ 868,
2128
+ 869,
2129
+ 870,
2130
+ 871,
2131
+ 872,
2132
+ 873,
2133
+ 874,
2134
+ 875,
2135
+ 876,
2136
+ 877,
2137
+ 878,
2138
+ 879,
2139
+ 880,
2140
+ 881,
2141
+ 882,
2142
+ 883,
2143
+ 884,
2144
+ 885,
2145
+ 886,
2146
+ 887,
2147
+ 888,
2148
+ 889,
2149
+ 890,
2150
+ 891,
2151
+ 892,
2152
+ 893,
2153
+ 894,
2154
+ 895,
2155
+ 896,
2156
+ 897,
2157
+ 898,
2158
+ 899,
2159
+ 900,
2160
+ 901,
2161
+ 902,
2162
+ 903,
2163
+ 904,
2164
+ 905,
2165
+ 906,
2166
+ 907,
2167
+ 908,
2168
+ 909,
2169
+ 910,
2170
+ 911,
2171
+ 912,
2172
+ 913,
2173
+ 914,
2174
+ 915,
2175
+ 916,
2176
+ 917,
2177
+ 918,
2178
+ 919,
2179
+ 920,
2180
+ 921,
2181
+ 922,
2182
+ 923,
2183
+ 924,
2184
+ 925,
2185
+ 926,
2186
+ 927,
2187
+ 928,
2188
+ 929,
2189
+ 930,
2190
+ 931,
2191
+ 932,
2192
+ 933,
2193
+ 934,
2194
+ 935,
2195
+ 936,
2196
+ 937,
2197
+ 938,
2198
+ 939,
2199
+ 940,
2200
+ 941,
2201
+ 942,
2202
+ 943,
2203
+ 944,
2204
+ 945,
2205
+ 946,
2206
+ 947,
2207
+ 948,
2208
+ 949,
2209
+ 950,
2210
+ 951,
2211
+ 952,
2212
+ 953,
2213
+ 954,
2214
+ 955,
2215
+ 956,
2216
+ 957,
2217
+ 958,
2218
+ 959,
2219
+ 960,
2220
+ 961,
2221
+ 962,
2222
+ 963,
2223
+ 964,
2224
+ 965,
2225
+ 966,
2226
+ 967,
2227
+ 968,
2228
+ 969,
2229
+ 970,
2230
+ 971,
2231
+ 972,
2232
+ 973,
2233
+ 974,
2234
+ 975,
2235
+ 976,
2236
+ 977,
2237
+ 978,
2238
+ 979,
2239
+ 980,
2240
+ 981,
2241
+ 982,
2242
+ 983,
2243
+ 984,
2244
+ 985,
2245
+ 986,
2246
+ 987,
2247
+ 988,
2248
+ 989,
2249
+ 990,
2250
+ 991,
2251
+ 992,
2252
+ 993,
2253
+ 994,
2254
+ 995,
2255
+ 996,
2256
+ 997,
2257
+ 998,
2258
+ 999,
2259
+ 1000,
2260
+ 1001,
2261
+ 1002,
2262
+ 1003,
2263
+ 1004,
2264
+ 1005,
2265
+ 1006,
2266
+ 1007,
2267
+ 1008,
2268
+ 1009,
2269
+ 1010,
2270
+ 1011,
2271
+ 1012,
2272
+ 1013,
2273
+ 1014,
2274
+ 1015,
2275
+ 1016,
2276
+ 1017,
2277
+ 1018,
2278
+ 1019,
2279
+ 1020,
2280
+ 1021,
2281
+ 1022,
2282
+ 1023
2283
+ ],
2284
+ "steps": null,
2285
+ "target_fps": 50.0,
2286
+ "valid_shape": [
2287
+ 1024,
2288
+ 199
2289
+ ]
2290
+ },
2291
+ "transition_metrics": {
2292
+ "chunk_length": 60,
2293
+ "enabled": false,
2294
+ "reason": "disabled"
2295
+ }
2296
+ },
2297
+ "label": "OmniMotion 300M mixed-modal scaling s050000"
2298
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/physical_metrics.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "body_jerk_mean": {
3
+ "max": 394.6324157714844,
4
+ "mean": 74.01262681838125,
5
+ "min": 10.021343231201172,
6
+ "std": 53.226579084617505
7
+ },
8
+ "contact_height_threshold": 0.12,
9
+ "contact_penetration_tolerance": 0.02,
10
+ "contact_sliding_speed": {
11
+ "max": 2.5723462104797363,
12
+ "mean": 0.35662794705444867,
13
+ "min": 0.0,
14
+ "std": 0.4447682159597098
15
+ },
16
+ "foot_ground_error": {
17
+ "max": 0.2917259931564331,
18
+ "mean": 0.04151647091404698,
19
+ "min": 0.00042374650365673006,
20
+ "std": 0.03340117411035358
21
+ },
22
+ "frames": 120,
23
+ "num_samples": 1024
24
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/real_motion_metrics.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "diversity_real": 1.3768246488138,
3
+ "matching_distance_real": {
4
+ "max": 1.4848475456237793,
5
+ "mean": 1.0944489734247327,
6
+ "min": 0.8405618071556091,
7
+ "std": 0.10523510496847523
8
+ },
9
+ "matching_score_real": 1.0944490432739258,
10
+ "motion_fid_reference_eval": 0.0,
11
+ "motion_fid_split_half": 0.0,
12
+ "motion_kid_reference_eval": {
13
+ "mean": -1.0749719189018236e-05,
14
+ "std": 0.0
15
+ },
16
+ "motion_kid_split_half": {
17
+ "mean": -1.0749719189018236e-05,
18
+ "std": 0.0
19
+ },
20
+ "multimodality": "-",
21
+ "num_reference_samples": 1024,
22
+ "num_samples": 1024,
23
+ "r_precision_real": [
24
+ 0.3740234375,
25
+ 0.5009765625,
26
+ 0.5810546875
27
+ ],
28
+ "text_real": {
29
+ "matching_distance": {
30
+ "max": 1.4848475456237793,
31
+ "mean": 1.0944489734247327,
32
+ "min": 0.8405618071556091,
33
+ "std": 0.10523510496847523
34
+ },
35
+ "matching_score": 1.0944490432739258,
36
+ "mean_text_rank": 5.232421875,
37
+ "median_text_rank": 2.0,
38
+ "r_precision": [
39
+ 0.3740234375,
40
+ 0.5009765625,
41
+ 0.5810546875
42
+ ],
43
+ "retrieval_batch_size": 32,
44
+ "retrieval_num_batches": 32,
45
+ "retrieval_num_samples": 1024,
46
+ "retrieval_order_policy": "manifest_order",
47
+ "retrieval_unique_datasets_per_batch": null
48
+ }
49
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/sample_metrics.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/sample_rankings.json ADDED
@@ -0,0 +1,1113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_generated": [
3
+ {
4
+ "caption": "sks person executes several throws with an object.",
5
+ "dataset": "permo_original_filtered_test",
6
+ "dataset_index": 1106,
7
+ "generated_matching_distance": 0.8289653062820435,
8
+ "generated_reference_distance_gap": -0.11820054054260254,
9
+ "generated_text_rank": 1,
10
+ "reference_matching_distance": 0.947165846824646,
11
+ "reference_text_rank": 1,
12
+ "sample_index": 896
13
+ },
14
+ {
15
+ "caption": "sks person propels themselves forward with several hops.",
16
+ "dataset": "permo_original_filtered_test",
17
+ "dataset_index": 735,
18
+ "generated_matching_distance": 0.8528786897659302,
19
+ "generated_reference_distance_gap": -0.053930044174194336,
20
+ "generated_text_rank": 1,
21
+ "reference_matching_distance": 0.9068087339401245,
22
+ "reference_text_rank": 1,
23
+ "sample_index": 878
24
+ },
25
+ {
26
+ "caption": "sks person makes several throwing motions, sending an item forward.",
27
+ "dataset": "permo_original_filtered_test",
28
+ "dataset_index": 1577,
29
+ "generated_matching_distance": 0.865251362323761,
30
+ "generated_reference_distance_gap": -0.03276187181472778,
31
+ "generated_text_rank": 1,
32
+ "reference_matching_distance": 0.8980132341384888,
33
+ "reference_text_rank": 1,
34
+ "sample_index": 927
35
+ },
36
+ {
37
+ "caption": "Punching forward several times, sks person uses one arm.",
38
+ "dataset": "permo_original_filtered_test",
39
+ "dataset_index": 1229,
40
+ "generated_matching_distance": 0.8728188276290894,
41
+ "generated_reference_distance_gap": -0.10825037956237793,
42
+ "generated_text_rank": 1,
43
+ "reference_matching_distance": 0.9810692071914673,
44
+ "reference_text_rank": 1,
45
+ "sample_index": 905
46
+ },
47
+ {
48
+ "caption": "sks person makes several throwing motions, sending an item forward.",
49
+ "dataset": "permo_original_filtered_test",
50
+ "dataset_index": 1579,
51
+ "generated_matching_distance": 0.8747169971466064,
52
+ "generated_reference_distance_gap": -0.026012301445007324,
53
+ "generated_text_rank": 1,
54
+ "reference_matching_distance": 0.9007292985916138,
55
+ "reference_text_rank": 1,
56
+ "sample_index": 928
57
+ },
58
+ {
59
+ "caption": "A person is standing still with hands clasped in prayer, radiating calmness and serenity.",
60
+ "dataset": "motiongv_original_filtered_test",
61
+ "dataset_index": 58580,
62
+ "generated_matching_distance": 0.8835017085075378,
63
+ "generated_reference_distance_gap": 0.00045418739318847656,
64
+ "generated_text_rank": 1,
65
+ "reference_matching_distance": 0.8830475211143494,
66
+ "reference_text_rank": 1,
67
+ "sample_index": 640
68
+ },
69
+ {
70
+ "caption": "Throwing multiple times, sks person propels an object through the air.",
71
+ "dataset": "permo_original_filtered_test",
72
+ "dataset_index": 810,
73
+ "generated_matching_distance": 0.8890764713287354,
74
+ "generated_reference_distance_gap": 0.03622448444366455,
75
+ "generated_text_rank": 1,
76
+ "reference_matching_distance": 0.8528519868850708,
77
+ "reference_text_rank": 1,
78
+ "sample_index": 881
79
+ },
80
+ {
81
+ "caption": "individual runs in place, bringing their knees up to meet their chin",
82
+ "dataset": "bones_seed_original_filtered_test",
83
+ "dataset_index": 8529,
84
+ "generated_matching_distance": 0.8937378525733948,
85
+ "generated_reference_distance_gap": -0.020252346992492676,
86
+ "generated_text_rank": 1,
87
+ "reference_matching_distance": 0.9139901995658875,
88
+ "reference_text_rank": 1,
89
+ "sample_index": 195
90
+ },
91
+ {
92
+ "caption": "sks person executes a series of kicks.",
93
+ "dataset": "permo_original_filtered_test",
94
+ "dataset_index": 1343,
95
+ "generated_matching_distance": 0.9017759561538696,
96
+ "generated_reference_distance_gap": 0.003324747085571289,
97
+ "generated_text_rank": 1,
98
+ "reference_matching_distance": 0.8984512090682983,
99
+ "reference_text_rank": 1,
100
+ "sample_index": 910
101
+ },
102
+ {
103
+ "caption": "A person uses their left hand to open and close the cabinet.",
104
+ "dataset": "motionllama_original_filtered_test",
105
+ "dataset_index": 1159,
106
+ "generated_matching_distance": 0.907444417476654,
107
+ "generated_reference_distance_gap": -0.05622255802154541,
108
+ "generated_text_rank": 1,
109
+ "reference_matching_distance": 0.9636669754981995,
110
+ "reference_text_rank": 1,
111
+ "sample_index": 749
112
+ },
113
+ {
114
+ "caption": "Someone walks forward with slight rotational steps, alternating legs, while moving its hands to mimic casting a fishing line and reeling it in, synchronizing the hand motions with its steps.; style: rhythmic",
115
+ "dataset": "idea400_original_filtered_test",
116
+ "dataset_index": 3107,
117
+ "generated_matching_distance": 0.9147061705589294,
118
+ "generated_reference_distance_gap": -0.032684147357940674,
119
+ "generated_text_rank": 1,
120
+ "reference_matching_distance": 0.9473903179168701,
121
+ "reference_text_rank": 1,
122
+ "sample_index": 512
123
+ },
124
+ {
125
+ "caption": "character turns 90 degrees right and proceeds to jog forward",
126
+ "dataset": "bones_seed_original_filtered_test",
127
+ "dataset_index": 22930,
128
+ "generated_matching_distance": 0.9150270819664001,
129
+ "generated_reference_distance_gap": -0.04880803823471069,
130
+ "generated_text_rank": 1,
131
+ "reference_matching_distance": 0.9638351202011108,
132
+ "reference_text_rank": 1,
133
+ "sample_index": 241
134
+ },
135
+ {
136
+ "caption": "A person uses the right hand to set down the phone and then uses the same hand to pick up the water bottle.",
137
+ "dataset": "motionllama_original_filtered_test",
138
+ "dataset_index": 1099,
139
+ "generated_matching_distance": 0.9240545630455017,
140
+ "generated_reference_distance_gap": -0.11129504442214966,
141
+ "generated_text_rank": 1,
142
+ "reference_matching_distance": 1.0353496074676514,
143
+ "reference_text_rank": 1,
144
+ "sample_index": 741
145
+ },
146
+ {
147
+ "caption": "a person jumps on the spot.",
148
+ "dataset": "humanml_original_filtered_test",
149
+ "dataset_index": 3101,
150
+ "generated_matching_distance": 0.9264796376228333,
151
+ "generated_reference_distance_gap": 0.04995793104171753,
152
+ "generated_text_rank": 1,
153
+ "reference_matching_distance": 0.8765217065811157,
154
+ "reference_text_rank": 1,
155
+ "sample_index": 382
156
+ },
157
+ {
158
+ "caption": "The person turns right and walks in a counterclockwise circle.",
159
+ "dataset": "snapmogen_original_filtered_test",
160
+ "dataset_index": 2071,
161
+ "generated_matching_distance": 0.9296107888221741,
162
+ "generated_reference_distance_gap": -0.06936514377593994,
163
+ "generated_text_rank": 1,
164
+ "reference_matching_distance": 0.998975932598114,
165
+ "reference_text_rank": 1,
166
+ "sample_index": 954
167
+ },
168
+ {
169
+ "caption": "The humanoid robot walks forward in a natural alternating step pattern, swinging each arm in opposition to the stepping foot for balance, and intermittently lifts one hand to the back of its head before returning the hand to its side while continuing to walk.; style: natural",
170
+ "dataset": "idea400_original_filtered_test",
171
+ "dataset_index": 3111,
172
+ "generated_matching_distance": 0.9325846433639526,
173
+ "generated_reference_distance_gap": 0.0013596415519714355,
174
+ "generated_text_rank": 1,
175
+ "reference_matching_distance": 0.9312250018119812,
176
+ "reference_text_rank": 1,
177
+ "sample_index": 513
178
+ },
179
+ {
180
+ "caption": "a person is boxing, jabbing mostly with his right hand.",
181
+ "dataset": "humanml_original_filtered_test",
182
+ "dataset_index": 3423,
183
+ "generated_matching_distance": 0.9330359697341919,
184
+ "generated_reference_distance_gap": 0.04328274726867676,
185
+ "generated_text_rank": 1,
186
+ "reference_matching_distance": 0.8897532224655151,
187
+ "reference_text_rank": 1,
188
+ "sample_index": 387
189
+ },
190
+ {
191
+ "caption": "The person crouches down, picks up something with his hands at foot level and lifts it while standing. Then he puts both feet together in jump, lifting something to his neck with his arms bent.",
192
+ "dataset": "snapmogen_original_filtered_test",
193
+ "dataset_index": 3972,
194
+ "generated_matching_distance": 0.9336654543876648,
195
+ "generated_reference_distance_gap": -0.06484144926071167,
196
+ "generated_text_rank": 1,
197
+ "reference_matching_distance": 0.9985069036483765,
198
+ "reference_text_rank": 1,
199
+ "sample_index": 964
200
+ },
201
+ {
202
+ "caption": "A woman demonstrates a hand gesture, emphasizing her right hand's movements with clarity and confidence.",
203
+ "dataset": "motiongv_original_filtered_test",
204
+ "dataset_index": 54151,
205
+ "generated_matching_distance": 0.9342674612998962,
206
+ "generated_reference_distance_gap": -0.004340648651123047,
207
+ "generated_text_rank": 1,
208
+ "reference_matching_distance": 0.9386081099510193,
209
+ "reference_text_rank": 1,
210
+ "sample_index": 637
211
+ },
212
+ {
213
+ "caption": "sks person kicks something on the ground multiple times.",
214
+ "dataset": "permo_original_filtered_test",
215
+ "dataset_index": 643,
216
+ "generated_matching_distance": 0.9351370334625244,
217
+ "generated_reference_distance_gap": -0.20336008071899414,
218
+ "generated_text_rank": 1,
219
+ "reference_matching_distance": 1.1384971141815186,
220
+ "reference_text_rank": 2,
221
+ "sample_index": 873
222
+ }
223
+ ],
224
+ "best_reference": [
225
+ {
226
+ "caption": "sks person performs a low kick.",
227
+ "dataset": "permo_original_filtered_test",
228
+ "dataset_index": 569,
229
+ "generated_matching_distance": 1.1705020666122437,
230
+ "generated_reference_distance_gap": 0.3299402594566345,
231
+ "generated_text_rank": 1,
232
+ "reference_matching_distance": 0.8405618071556091,
233
+ "reference_text_rank": 1,
234
+ "sample_index": 869
235
+ },
236
+ {
237
+ "caption": "Throwing multiple times, sks person propels an object through the air.",
238
+ "dataset": "permo_original_filtered_test",
239
+ "dataset_index": 810,
240
+ "generated_matching_distance": 0.8890764713287354,
241
+ "generated_reference_distance_gap": 0.03622448444366455,
242
+ "generated_text_rank": 1,
243
+ "reference_matching_distance": 0.8528519868850708,
244
+ "reference_text_rank": 1,
245
+ "sample_index": 881
246
+ },
247
+ {
248
+ "caption": "walking forward with hands placed behind the back",
249
+ "dataset": "bones_seed_original_filtered_test",
250
+ "dataset_index": 25768,
251
+ "generated_matching_distance": 0.9673618078231812,
252
+ "generated_reference_distance_gap": 0.10176974534988403,
253
+ "generated_text_rank": 1,
254
+ "reference_matching_distance": 0.8655920624732971,
255
+ "reference_text_rank": 1,
256
+ "sample_index": 252
257
+ },
258
+ {
259
+ "caption": "sks person is bouncing ahead on one foot.",
260
+ "dataset": "permo_original_filtered_test",
261
+ "dataset_index": 433,
262
+ "generated_matching_distance": 0.9449302554130554,
263
+ "generated_reference_distance_gap": 0.06853574514389038,
264
+ "generated_text_rank": 1,
265
+ "reference_matching_distance": 0.876394510269165,
266
+ "reference_text_rank": 1,
267
+ "sample_index": 863
268
+ },
269
+ {
270
+ "caption": "a person jumps on the spot.",
271
+ "dataset": "humanml_original_filtered_test",
272
+ "dataset_index": 3101,
273
+ "generated_matching_distance": 0.9264796376228333,
274
+ "generated_reference_distance_gap": 0.04995793104171753,
275
+ "generated_text_rank": 1,
276
+ "reference_matching_distance": 0.8765217065811157,
277
+ "reference_text_rank": 1,
278
+ "sample_index": 382
279
+ },
280
+ {
281
+ "caption": "figure covering their mouth with hands",
282
+ "dataset": "bones_seed_original_filtered_test",
283
+ "dataset_index": 18940,
284
+ "generated_matching_distance": 1.1072214841842651,
285
+ "generated_reference_distance_gap": 0.22756928205490112,
286
+ "generated_text_rank": 1,
287
+ "reference_matching_distance": 0.879652202129364,
288
+ "reference_text_rank": 1,
289
+ "sample_index": 226
290
+ },
291
+ {
292
+ "caption": "A person is doing the Dumbbell Curl Trifecta exercise.",
293
+ "dataset": "motionllama_original_filtered_test",
294
+ "dataset_index": 296,
295
+ "generated_matching_distance": 1.2189940214157104,
296
+ "generated_reference_distance_gap": 0.33789384365081787,
297
+ "generated_text_rank": 1,
298
+ "reference_matching_distance": 0.8811001777648926,
299
+ "reference_text_rank": 1,
300
+ "sample_index": 700
301
+ },
302
+ {
303
+ "caption": "a person side steps to the left.",
304
+ "dataset": "humanml_original_filtered_test",
305
+ "dataset_index": 4879,
306
+ "generated_matching_distance": 1.0454336404800415,
307
+ "generated_reference_distance_gap": 0.16327708959579468,
308
+ "generated_text_rank": 1,
309
+ "reference_matching_distance": 0.8821565508842468,
310
+ "reference_text_rank": 1,
311
+ "sample_index": 403
312
+ },
313
+ {
314
+ "caption": "The person walks forward. Then crouches and jumps forward, turning right in mid-air, stands up straight, turns right and walks straight, crouches and jumps forward turning right in mid-air and stands up straight.",
315
+ "dataset": "snapmogen_original_filtered_test",
316
+ "dataset_index": 7470,
317
+ "generated_matching_distance": 1.006026268005371,
318
+ "generated_reference_distance_gap": 0.12371188402175903,
319
+ "generated_text_rank": 1,
320
+ "reference_matching_distance": 0.8823143839836121,
321
+ "reference_text_rank": 1,
322
+ "sample_index": 984
323
+ },
324
+ {
325
+ "caption": "A person is standing still with hands clasped in prayer, radiating calmness and serenity.",
326
+ "dataset": "motiongv_original_filtered_test",
327
+ "dataset_index": 58580,
328
+ "generated_matching_distance": 0.8835017085075378,
329
+ "generated_reference_distance_gap": 0.00045418739318847656,
330
+ "generated_text_rank": 1,
331
+ "reference_matching_distance": 0.8830475211143494,
332
+ "reference_text_rank": 1,
333
+ "sample_index": 640
334
+ },
335
+ {
336
+ "caption": "sks person hops forward on one leg, repeating the motion several times.",
337
+ "dataset": "permo_original_filtered_test",
338
+ "dataset_index": 236,
339
+ "generated_matching_distance": 1.0698020458221436,
340
+ "generated_reference_distance_gap": 0.18311238288879395,
341
+ "generated_text_rank": 2,
342
+ "reference_matching_distance": 0.8866896629333496,
343
+ "reference_text_rank": 1,
344
+ "sample_index": 859
345
+ },
346
+ {
347
+ "caption": "a person is boxing, jabbing mostly with his right hand.",
348
+ "dataset": "humanml_original_filtered_test",
349
+ "dataset_index": 3423,
350
+ "generated_matching_distance": 0.9330359697341919,
351
+ "generated_reference_distance_gap": 0.04328274726867676,
352
+ "generated_text_rank": 1,
353
+ "reference_matching_distance": 0.8897532224655151,
354
+ "reference_text_rank": 1,
355
+ "sample_index": 387
356
+ },
357
+ {
358
+ "caption": "a person walks backwards slowly.",
359
+ "dataset": "humanml_original_filtered_test",
360
+ "dataset_index": 3058,
361
+ "generated_matching_distance": 1.2436095476150513,
362
+ "generated_reference_distance_gap": 0.3536805510520935,
363
+ "generated_text_rank": 6,
364
+ "reference_matching_distance": 0.8899289965629578,
365
+ "reference_text_rank": 1,
366
+ "sample_index": 381
367
+ },
368
+ {
369
+ "caption": "sks takes several forward jumps.",
370
+ "dataset": "permo_original_filtered_test",
371
+ "dataset_index": 944,
372
+ "generated_matching_distance": 1.0714632272720337,
373
+ "generated_reference_distance_gap": 0.17660194635391235,
374
+ "generated_text_rank": 1,
375
+ "reference_matching_distance": 0.8948612809181213,
376
+ "reference_text_rank": 1,
377
+ "sample_index": 888
378
+ },
379
+ {
380
+ "caption": "sks punches straight ahead with one arm several times.",
381
+ "dataset": "permo_original_filtered_test",
382
+ "dataset_index": 1519,
383
+ "generated_matching_distance": 1.2389142513275146,
384
+ "generated_reference_distance_gap": 0.34369730949401855,
385
+ "generated_text_rank": 5,
386
+ "reference_matching_distance": 0.8952169418334961,
387
+ "reference_text_rank": 1,
388
+ "sample_index": 920
389
+ },
390
+ {
391
+ "caption": "sks person targets an object on the floor with several kicks.",
392
+ "dataset": "permo_original_filtered_test",
393
+ "dataset_index": 6,
394
+ "generated_matching_distance": 1.126343011856079,
395
+ "generated_reference_distance_gap": 0.23065459728240967,
396
+ "generated_text_rank": 1,
397
+ "reference_matching_distance": 0.8956884145736694,
398
+ "reference_text_rank": 1,
399
+ "sample_index": 855
400
+ },
401
+ {
402
+ "caption": "A man walks forward, alternating between legs with each step while flexing its knees to bring its heels toward its buttocks.; style: rhythmic",
403
+ "dataset": "idea400_original_filtered_test",
404
+ "dataset_index": 1568,
405
+ "generated_matching_distance": 1.0014300346374512,
406
+ "generated_reference_distance_gap": 0.10570824146270752,
407
+ "generated_text_rank": 1,
408
+ "reference_matching_distance": 0.8957217931747437,
409
+ "reference_text_rank": 1,
410
+ "sample_index": 468
411
+ },
412
+ {
413
+ "caption": "character begins to stand with their hands behind their back",
414
+ "dataset": "bones_seed_original_filtered_test",
415
+ "dataset_index": 9657,
416
+ "generated_matching_distance": 1.305927038192749,
417
+ "generated_reference_distance_gap": 0.40971076488494873,
418
+ "generated_text_rank": 4,
419
+ "reference_matching_distance": 0.8962162733078003,
420
+ "reference_text_rank": 1,
421
+ "sample_index": 196
422
+ },
423
+ {
424
+ "caption": "sks person waves to someone, moving one arm repeatedly.",
425
+ "dataset": "permo_original_filtered_test",
426
+ "dataset_index": 489,
427
+ "generated_matching_distance": 1.007538914680481,
428
+ "generated_reference_distance_gap": 0.11073499917984009,
429
+ "generated_text_rank": 1,
430
+ "reference_matching_distance": 0.8968039155006409,
431
+ "reference_text_rank": 1,
432
+ "sample_index": 866
433
+ },
434
+ {
435
+ "caption": "sks person makes several throwing motions, sending an item forward.",
436
+ "dataset": "permo_original_filtered_test",
437
+ "dataset_index": 1577,
438
+ "generated_matching_distance": 0.865251362323761,
439
+ "generated_reference_distance_gap": -0.03276187181472778,
440
+ "generated_text_rank": 1,
441
+ "reference_matching_distance": 0.8980132341384888,
442
+ "reference_text_rank": 1,
443
+ "sample_index": 927
444
+ }
445
+ ],
446
+ "largest_generated_reference_distance_gap": [
447
+ {
448
+ "caption": "character begins to stand with their hands behind their back",
449
+ "dataset": "bones_seed_original_filtered_test",
450
+ "dataset_index": 9657,
451
+ "generated_matching_distance": 1.305927038192749,
452
+ "generated_reference_distance_gap": 0.40971076488494873,
453
+ "generated_text_rank": 4,
454
+ "reference_matching_distance": 0.8962162733078003,
455
+ "reference_text_rank": 1,
456
+ "sample_index": 196
457
+ },
458
+ {
459
+ "caption": "sks person jumps forward several times.",
460
+ "dataset": "permo_original_filtered_test",
461
+ "dataset_index": 913,
462
+ "generated_matching_distance": 1.3080419301986694,
463
+ "generated_reference_distance_gap": 0.40285414457321167,
464
+ "generated_text_rank": 6,
465
+ "reference_matching_distance": 0.9051877856254578,
466
+ "reference_text_rank": 1,
467
+ "sample_index": 884
468
+ },
469
+ {
470
+ "caption": "sks is walking forward, then turning to the side with a pivot.",
471
+ "dataset": "permo_original_filtered_test",
472
+ "dataset_index": 1534,
473
+ "generated_matching_distance": 1.412964940071106,
474
+ "generated_reference_distance_gap": 0.39798033237457275,
475
+ "generated_text_rank": 23,
476
+ "reference_matching_distance": 1.0149846076965332,
477
+ "reference_text_rank": 2,
478
+ "sample_index": 923
479
+ },
480
+ {
481
+ "caption": "A person uses his right hand to put down the keyboard, open the drawer, pick up the mouse, and then put it down.",
482
+ "dataset": "motionllama_original_filtered_test",
483
+ "dataset_index": 1111,
484
+ "generated_matching_distance": 1.3164666891098022,
485
+ "generated_reference_distance_gap": 0.3698711395263672,
486
+ "generated_text_rank": 7,
487
+ "reference_matching_distance": 0.9465955495834351,
488
+ "reference_text_rank": 1,
489
+ "sample_index": 742
490
+ },
491
+ {
492
+ "caption": "sks waves with one arm, making several motions.",
493
+ "dataset": "permo_original_filtered_test",
494
+ "dataset_index": 1482,
495
+ "generated_matching_distance": 1.338932752609253,
496
+ "generated_reference_distance_gap": 0.3674030900001526,
497
+ "generated_text_rank": 5,
498
+ "reference_matching_distance": 0.9715296626091003,
499
+ "reference_text_rank": 1,
500
+ "sample_index": 918
501
+ },
502
+ {
503
+ "caption": "The person stands, wearily bringing their right hand to their mouth, leaning their torso forward, and nodding their head from side to side while lowering their right hand. They then raise both hands to their cheeks, shaking them near their face.",
504
+ "dataset": "snapmogen_original_filtered_test",
505
+ "dataset_index": 11536,
506
+ "generated_matching_distance": 1.2637330293655396,
507
+ "generated_reference_distance_gap": 0.3624747395515442,
508
+ "generated_text_rank": 1,
509
+ "reference_matching_distance": 0.9012582898139954,
510
+ "reference_text_rank": 1,
511
+ "sample_index": 1021
512
+ },
513
+ {
514
+ "caption": "The person raises the right arm up and straightens it to the right, simultaneously straightening the left arm to the left. Then he rotates both arms, stretching the shoulders. After he tilts the body forward and takes both arms back, brings the arms forward and raises them. Then he spreads them apart and brings them down.",
515
+ "dataset": "snapmogen_original_filtered_test",
516
+ "dataset_index": 6766,
517
+ "generated_matching_distance": 1.3117051124572754,
518
+ "generated_reference_distance_gap": 0.3554818630218506,
519
+ "generated_text_rank": 4,
520
+ "reference_matching_distance": 0.9562232494354248,
521
+ "reference_text_rank": 1,
522
+ "sample_index": 980
523
+ },
524
+ {
525
+ "caption": "a person walks backwards slowly.",
526
+ "dataset": "humanml_original_filtered_test",
527
+ "dataset_index": 3058,
528
+ "generated_matching_distance": 1.2436095476150513,
529
+ "generated_reference_distance_gap": 0.3536805510520935,
530
+ "generated_text_rank": 6,
531
+ "reference_matching_distance": 0.8899289965629578,
532
+ "reference_text_rank": 1,
533
+ "sample_index": 381
534
+ },
535
+ {
536
+ "caption": "character sorts out small heavy object with two hands from low up front and puts it away low behind",
537
+ "dataset": "bones_seed_original_filtered_test",
538
+ "dataset_index": 21877,
539
+ "generated_matching_distance": 1.2763309478759766,
540
+ "generated_reference_distance_gap": 0.34587401151657104,
541
+ "generated_text_rank": 3,
542
+ "reference_matching_distance": 0.9304569363594055,
543
+ "reference_text_rank": 1,
544
+ "sample_index": 240
545
+ },
546
+ {
547
+ "caption": "sks punches straight ahead with one arm several times.",
548
+ "dataset": "permo_original_filtered_test",
549
+ "dataset_index": 1519,
550
+ "generated_matching_distance": 1.2389142513275146,
551
+ "generated_reference_distance_gap": 0.34369730949401855,
552
+ "generated_text_rank": 5,
553
+ "reference_matching_distance": 0.8952169418334961,
554
+ "reference_text_rank": 1,
555
+ "sample_index": 920
556
+ },
557
+ {
558
+ "caption": "A person is doing the Dumbbell Curl Trifecta exercise.",
559
+ "dataset": "motionllama_original_filtered_test",
560
+ "dataset_index": 296,
561
+ "generated_matching_distance": 1.2189940214157104,
562
+ "generated_reference_distance_gap": 0.33789384365081787,
563
+ "generated_text_rank": 1,
564
+ "reference_matching_distance": 0.8811001777648926,
565
+ "reference_text_rank": 1,
566
+ "sample_index": 700
567
+ },
568
+ {
569
+ "caption": "sks person hops forward multiple times.",
570
+ "dataset": "permo_original_filtered_test",
571
+ "dataset_index": 933,
572
+ "generated_matching_distance": 1.3304530382156372,
573
+ "generated_reference_distance_gap": 0.3305116295814514,
574
+ "generated_text_rank": 14,
575
+ "reference_matching_distance": 0.9999414086341858,
576
+ "reference_text_rank": 2,
577
+ "sample_index": 886
578
+ },
579
+ {
580
+ "caption": "sks person performs a low kick.",
581
+ "dataset": "permo_original_filtered_test",
582
+ "dataset_index": 569,
583
+ "generated_matching_distance": 1.1705020666122437,
584
+ "generated_reference_distance_gap": 0.3299402594566345,
585
+ "generated_text_rank": 1,
586
+ "reference_matching_distance": 0.8405618071556091,
587
+ "reference_text_rank": 1,
588
+ "sample_index": 869
589
+ },
590
+ {
591
+ "caption": "person walks forwards then takes a large step over something",
592
+ "dataset": "humanml_original_filtered_test",
593
+ "dataset_index": 3879,
594
+ "generated_matching_distance": 1.320884108543396,
595
+ "generated_reference_distance_gap": 0.3267359733581543,
596
+ "generated_text_rank": 12,
597
+ "reference_matching_distance": 0.9941481351852417,
598
+ "reference_text_rank": 1,
599
+ "sample_index": 390
600
+ },
601
+ {
602
+ "caption": "A person uses their right hand to pick up the phone, their left hand to put it down, and then sits.",
603
+ "dataset": "motionllama_original_filtered_test",
604
+ "dataset_index": 1081,
605
+ "generated_matching_distance": 1.3333842754364014,
606
+ "generated_reference_distance_gap": 0.3246074914932251,
607
+ "generated_text_rank": 9,
608
+ "reference_matching_distance": 1.0087767839431763,
609
+ "reference_text_rank": 2,
610
+ "sample_index": 740
611
+ },
612
+ {
613
+ "caption": "right-handed person playing violin",
614
+ "dataset": "bones_seed_original_filtered_test",
615
+ "dataset_index": 19933,
616
+ "generated_matching_distance": 1.3572005033493042,
617
+ "generated_reference_distance_gap": 0.3221616744995117,
618
+ "generated_text_rank": 7,
619
+ "reference_matching_distance": 1.0350388288497925,
620
+ "reference_text_rank": 1,
621
+ "sample_index": 234
622
+ },
623
+ {
624
+ "caption": "He assumes a stable stance, with feet shoulder-width apart and core engaged, and prepares for side lateral raises by positioning his hands at his sides. He then raises both arms to the sides, keeping them straight until parallel to the ground, while maintaining a neutral shoulder position and focusing on deltoid contraction. At the peak of the movement, his arms form a \"T\" shape, and he briefly pauses to experience the full contraction of his deltoids. Next, he slowly lowers his arms back to the starting position, maintaining control throughout the descent. This completes one full side lateral raise, which he repeats three times.",
625
+ "dataset": "motionllama_original_filtered_test",
626
+ "dataset_index": 394,
627
+ "generated_matching_distance": 1.2824586629867554,
628
+ "generated_reference_distance_gap": 0.31979262828826904,
629
+ "generated_text_rank": 3,
630
+ "reference_matching_distance": 0.9626660346984863,
631
+ "reference_text_rank": 1,
632
+ "sample_index": 706
633
+ },
634
+ {
635
+ "caption": "They are doing dumbbell bicep curls.",
636
+ "dataset": "motionllama_original_filtered_test",
637
+ "dataset_index": 289,
638
+ "generated_matching_distance": 1.2388211488723755,
639
+ "generated_reference_distance_gap": 0.3155640959739685,
640
+ "generated_text_rank": 1,
641
+ "reference_matching_distance": 0.923257052898407,
642
+ "reference_text_rank": 1,
643
+ "sample_index": 699
644
+ },
645
+ {
646
+ "caption": "the person is turning on the shower with her right arm.",
647
+ "dataset": "humanml_original_filtered_test",
648
+ "dataset_index": 7179,
649
+ "generated_matching_distance": 1.4400562047958374,
650
+ "generated_reference_distance_gap": 0.30823826789855957,
651
+ "generated_text_rank": 25,
652
+ "reference_matching_distance": 1.1318179368972778,
653
+ "reference_text_rank": 1,
654
+ "sample_index": 424
655
+ },
656
+ {
657
+ "caption": "sks makes several kicks towards an object on the floor.",
658
+ "dataset": "permo_original_filtered_test",
659
+ "dataset_index": 1336,
660
+ "generated_matching_distance": 1.2394676208496094,
661
+ "generated_reference_distance_gap": 0.3062068223953247,
662
+ "generated_text_rank": 2,
663
+ "reference_matching_distance": 0.9332607984542847,
664
+ "reference_text_rank": 1,
665
+ "sample_index": 909
666
+ }
667
+ ],
668
+ "num_samples": 1024,
669
+ "worst_generated": [
670
+ {
671
+ "caption": "stand still, then bend forward with hands on knees; style: calm",
672
+ "dataset": "lafan1_original_filtered_test",
673
+ "dataset_index": 303,
674
+ "generated_matching_distance": 1.446708083152771,
675
+ "generated_reference_distance_gap": 0.028352856636047363,
676
+ "generated_text_rank": 31,
677
+ "reference_matching_distance": 1.4183552265167236,
678
+ "reference_text_rank": 19,
679
+ "sample_index": 580
680
+ },
681
+ {
682
+ "caption": "step into wide bent-knee stance, hold position, then return to narrow stance; style: dynamic",
683
+ "dataset": "amass_original_filtered_test",
684
+ "dataset_index": 8644,
685
+ "generated_matching_distance": 1.5543382167816162,
686
+ "generated_reference_distance_gap": 0.06949067115783691,
687
+ "generated_text_rank": 29,
688
+ "reference_matching_distance": 1.4848475456237793,
689
+ "reference_text_rank": 20,
690
+ "sample_index": 166
691
+ },
692
+ {
693
+ "caption": "stand upright with arms extended horizontally; style: static",
694
+ "dataset": "lafan1_original_filtered_test",
695
+ "dataset_index": 355,
696
+ "generated_matching_distance": 1.44863760471344,
697
+ "generated_reference_distance_gap": 0.12539052963256836,
698
+ "generated_text_rank": 29,
699
+ "reference_matching_distance": 1.3232470750808716,
700
+ "reference_text_rank": 13,
701
+ "sample_index": 589
702
+ },
703
+ {
704
+ "caption": "The humanoid robot moves forward with one leg, bends knees slightly while halting, raises one arm forward in a stopping gesture, returns the arm to its side, then assumes a stable standing position with feet shoulder-width apart.; style: controlled",
705
+ "dataset": "idea400_original_filtered_test",
706
+ "dataset_index": 1770,
707
+ "generated_matching_distance": 1.4350745677947998,
708
+ "generated_reference_distance_gap": 0.1551605463027954,
709
+ "generated_text_rank": 28,
710
+ "reference_matching_distance": 1.2799140214920044,
711
+ "reference_text_rank": 6,
712
+ "sample_index": 472
713
+ },
714
+ {
715
+ "caption": "lower arms and return to neutral standing position; style: calm",
716
+ "dataset": "omomo_original_filtered_test",
717
+ "dataset_index": 1253,
718
+ "generated_matching_distance": 1.4282751083374023,
719
+ "generated_reference_distance_gap": 0.1728992462158203,
720
+ "generated_text_rank": 26,
721
+ "reference_matching_distance": 1.255375862121582,
722
+ "reference_text_rank": 4,
723
+ "sample_index": 812
724
+ },
725
+ {
726
+ "caption": "the person is turning on the shower with her right arm.",
727
+ "dataset": "humanml_original_filtered_test",
728
+ "dataset_index": 7179,
729
+ "generated_matching_distance": 1.4400562047958374,
730
+ "generated_reference_distance_gap": 0.30823826789855957,
731
+ "generated_text_rank": 25,
732
+ "reference_matching_distance": 1.1318179368972778,
733
+ "reference_text_rank": 1,
734
+ "sample_index": 424
735
+ },
736
+ {
737
+ "caption": "repeat squat down and return to standing motion; style: controlled",
738
+ "dataset": "amass_original_filtered_test",
739
+ "dataset_index": 8679,
740
+ "generated_matching_distance": 1.4264090061187744,
741
+ "generated_reference_distance_gap": 0.1630847454071045,
742
+ "generated_text_rank": 23,
743
+ "reference_matching_distance": 1.26332426071167,
744
+ "reference_text_rank": 2,
745
+ "sample_index": 169
746
+ },
747
+ {
748
+ "caption": "bend torso forward and extend arms downward; style: controlled",
749
+ "dataset": "omomo_original_filtered_test",
750
+ "dataset_index": 1959,
751
+ "generated_matching_distance": 1.419572114944458,
752
+ "generated_reference_distance_gap": 0.19353985786437988,
753
+ "generated_text_rank": 23,
754
+ "reference_matching_distance": 1.2260322570800781,
755
+ "reference_text_rank": 5,
756
+ "sample_index": 831
757
+ },
758
+ {
759
+ "caption": "sks is walking forward, then turning to the side with a pivot.",
760
+ "dataset": "permo_original_filtered_test",
761
+ "dataset_index": 1534,
762
+ "generated_matching_distance": 1.412964940071106,
763
+ "generated_reference_distance_gap": 0.39798033237457275,
764
+ "generated_text_rank": 23,
765
+ "reference_matching_distance": 1.0149846076965332,
766
+ "reference_text_rank": 2,
767
+ "sample_index": 923
768
+ },
769
+ {
770
+ "caption": "rotate full circle with arms extended outward; style: smooth",
771
+ "dataset": "omomo_original_filtered_test",
772
+ "dataset_index": 1704,
773
+ "generated_matching_distance": 1.3914982080459595,
774
+ "generated_reference_distance_gap": 0.18154561519622803,
775
+ "generated_text_rank": 21,
776
+ "reference_matching_distance": 1.2099525928497314,
777
+ "reference_text_rank": 2,
778
+ "sample_index": 822
779
+ },
780
+ {
781
+ "caption": "Happily, sport fitness enthusiasts perform a standing left and right arm stretch.",
782
+ "dataset": "fitness_original_filtered_test",
783
+ "dataset_index": 75,
784
+ "generated_matching_distance": 1.3753480911254883,
785
+ "generated_reference_distance_gap": 0.10955023765563965,
786
+ "generated_text_rank": 21,
787
+ "reference_matching_distance": 1.2657978534698486,
788
+ "reference_text_rank": 4,
789
+ "sample_index": 317
790
+ },
791
+ {
792
+ "caption": "rotate to face away, bend torso forward, then straighten up; style: fluid",
793
+ "dataset": "omomo_original_filtered_test",
794
+ "dataset_index": 930,
795
+ "generated_matching_distance": 1.4157949686050415,
796
+ "generated_reference_distance_gap": 0.18482637405395508,
797
+ "generated_text_rank": 20,
798
+ "reference_matching_distance": 1.2309685945510864,
799
+ "reference_text_rank": 4,
800
+ "sample_index": 800
801
+ },
802
+ {
803
+ "caption": "A man performs a stretching exercise, standing shoulder-width apart with a band held in both hands, raising it above his head, bending to the side, and repeating the movement with focus in a well-lit space.",
804
+ "dataset": "motiongv_original_filtered_test",
805
+ "dataset_index": 47711,
806
+ "generated_matching_distance": 1.4481498003005981,
807
+ "generated_reference_distance_gap": 0.24703335762023926,
808
+ "generated_text_rank": 18,
809
+ "reference_matching_distance": 1.2011164426803589,
810
+ "reference_text_rank": 1,
811
+ "sample_index": 631
812
+ },
813
+ {
814
+ "caption": "sks person makes a transition.",
815
+ "dataset": "permo_original_filtered_test",
816
+ "dataset_index": 1464,
817
+ "generated_matching_distance": 1.3064711093902588,
818
+ "generated_reference_distance_gap": 0.19771075248718262,
819
+ "generated_text_rank": 18,
820
+ "reference_matching_distance": 1.1087603569030762,
821
+ "reference_text_rank": 2,
822
+ "sample_index": 917
823
+ },
824
+ {
825
+ "caption": "A person places the handbag with their right hand.",
826
+ "dataset": "motionllama_original_filtered_test",
827
+ "dataset_index": 1234,
828
+ "generated_matching_distance": 1.338201642036438,
829
+ "generated_reference_distance_gap": 0.15164148807525635,
830
+ "generated_text_rank": 16,
831
+ "reference_matching_distance": 1.1865601539611816,
832
+ "reference_text_rank": 2,
833
+ "sample_index": 754
834
+ },
835
+ {
836
+ "caption": "turn to face forward, wave arms circularly then lower; style: expressive",
837
+ "dataset": "omomo_original_filtered_test",
838
+ "dataset_index": 1436,
839
+ "generated_matching_distance": 1.3788769245147705,
840
+ "generated_reference_distance_gap": 0.14061760902404785,
841
+ "generated_text_rank": 15,
842
+ "reference_matching_distance": 1.2382593154907227,
843
+ "reference_text_rank": 4,
844
+ "sample_index": 814
845
+ },
846
+ {
847
+ "caption": "step, raise arm, then return to T-pose; style: concluding",
848
+ "dataset": "lafan1_original_filtered_test",
849
+ "dataset_index": 78,
850
+ "generated_matching_distance": 1.3633464574813843,
851
+ "generated_reference_distance_gap": 0.12143599987030029,
852
+ "generated_text_rank": 15,
853
+ "reference_matching_distance": 1.241910457611084,
854
+ "reference_text_rank": 1,
855
+ "sample_index": 529
856
+ },
857
+ {
858
+ "caption": "wave hands up/down, then lower arms gradually; style: controlled",
859
+ "dataset": "omomo_original_filtered_test",
860
+ "dataset_index": 2826,
861
+ "generated_matching_distance": 1.398184895515442,
862
+ "generated_reference_distance_gap": 0.14275777339935303,
863
+ "generated_text_rank": 14,
864
+ "reference_matching_distance": 1.2554271221160889,
865
+ "reference_text_rank": 5,
866
+ "sample_index": 852
867
+ },
868
+ {
869
+ "caption": "sks person hops forward multiple times.",
870
+ "dataset": "permo_original_filtered_test",
871
+ "dataset_index": 933,
872
+ "generated_matching_distance": 1.3304530382156372,
873
+ "generated_reference_distance_gap": 0.3305116295814514,
874
+ "generated_text_rank": 14,
875
+ "reference_matching_distance": 0.9999414086341858,
876
+ "reference_text_rank": 2,
877
+ "sample_index": 886
878
+ },
879
+ {
880
+ "caption": "lower body into squat position and push back up to standing; style: controlled",
881
+ "dataset": "amass_original_filtered_test",
882
+ "dataset_index": 8677,
883
+ "generated_matching_distance": 1.3198378086090088,
884
+ "generated_reference_distance_gap": 0.051113128662109375,
885
+ "generated_text_rank": 14,
886
+ "reference_matching_distance": 1.2687246799468994,
887
+ "reference_text_rank": 2,
888
+ "sample_index": 168
889
+ }
890
+ ],
891
+ "worst_reference": [
892
+ {
893
+ "caption": "step into wide bent-knee stance, hold position, then return to narrow stance; style: dynamic",
894
+ "dataset": "amass_original_filtered_test",
895
+ "dataset_index": 8644,
896
+ "generated_matching_distance": 1.5543382167816162,
897
+ "generated_reference_distance_gap": 0.06949067115783691,
898
+ "generated_text_rank": 29,
899
+ "reference_matching_distance": 1.4848475456237793,
900
+ "reference_text_rank": 20,
901
+ "sample_index": 166
902
+ },
903
+ {
904
+ "caption": "stand still, then bend forward with hands on knees; style: calm",
905
+ "dataset": "lafan1_original_filtered_test",
906
+ "dataset_index": 303,
907
+ "generated_matching_distance": 1.446708083152771,
908
+ "generated_reference_distance_gap": 0.028352856636047363,
909
+ "generated_text_rank": 31,
910
+ "reference_matching_distance": 1.4183552265167236,
911
+ "reference_text_rank": 19,
912
+ "sample_index": 580
913
+ },
914
+ {
915
+ "caption": "perform repetitive side steps and arm movements; style: rhythmic",
916
+ "dataset": "amass_original_filtered_test",
917
+ "dataset_index": 8230,
918
+ "generated_matching_distance": 1.257988691329956,
919
+ "generated_reference_distance_gap": -0.07763147354125977,
920
+ "generated_text_rank": 6,
921
+ "reference_matching_distance": 1.3356201648712158,
922
+ "reference_text_rank": 19,
923
+ "sample_index": 157
924
+ },
925
+ {
926
+ "caption": "The humanoid robot performs an in-place dance routine, shifting weight between feet, slightly bending knees, and moving arms in coordination with leg motions.; style: rhythmic",
927
+ "dataset": "idea400_original_filtered_test",
928
+ "dataset_index": 204,
929
+ "generated_matching_distance": 1.0607038736343384,
930
+ "generated_reference_distance_gap": -0.33269202709198,
931
+ "generated_text_rank": 1,
932
+ "reference_matching_distance": 1.3933959007263184,
933
+ "reference_text_rank": 14,
934
+ "sample_index": 433
935
+ },
936
+ {
937
+ "caption": "alternate leg lifts with quick arm movements; style: energetic",
938
+ "dataset": "100style_original_filtered_test",
939
+ "dataset_index": 1869,
940
+ "generated_matching_distance": 1.308235764503479,
941
+ "generated_reference_distance_gap": -0.07149863243103027,
942
+ "generated_text_rank": 11,
943
+ "reference_matching_distance": 1.3797343969345093,
944
+ "reference_text_rank": 14,
945
+ "sample_index": 40
946
+ },
947
+ {
948
+ "caption": "walk forward with alternating leg movements and gentle arm swings; style: rhythmic, controlled",
949
+ "dataset": "amass_original_filtered_test",
950
+ "dataset_index": 1023,
951
+ "generated_matching_distance": 1.161105990409851,
952
+ "generated_reference_distance_gap": -0.15137624740600586,
953
+ "generated_text_rank": 1,
954
+ "reference_matching_distance": 1.312482237815857,
955
+ "reference_text_rank": 14,
956
+ "sample_index": 94
957
+ },
958
+ {
959
+ "caption": "stand upright with arms extended horizontally; style: static",
960
+ "dataset": "lafan1_original_filtered_test",
961
+ "dataset_index": 355,
962
+ "generated_matching_distance": 1.44863760471344,
963
+ "generated_reference_distance_gap": 0.12539052963256836,
964
+ "generated_text_rank": 29,
965
+ "reference_matching_distance": 1.3232470750808716,
966
+ "reference_text_rank": 13,
967
+ "sample_index": 589
968
+ },
969
+ {
970
+ "caption": "stand still facing backward; style: static",
971
+ "dataset": "amass_original_filtered_test",
972
+ "dataset_index": 7974,
973
+ "generated_matching_distance": 1.2389743328094482,
974
+ "generated_reference_distance_gap": -0.027637600898742676,
975
+ "generated_text_rank": 6,
976
+ "reference_matching_distance": 1.266611933708191,
977
+ "reference_text_rank": 12,
978
+ "sample_index": 154
979
+ },
980
+ {
981
+ "caption": "A man in athletic wear is sprinting forward, mimicking a marathon runner's powerful stride.",
982
+ "dataset": "motiongv_original_filtered_test",
983
+ "dataset_index": 63912,
984
+ "generated_matching_distance": 1.2016425132751465,
985
+ "generated_reference_distance_gap": -0.1641627550125122,
986
+ "generated_text_rank": 5,
987
+ "reference_matching_distance": 1.3658052682876587,
988
+ "reference_text_rank": 11,
989
+ "sample_index": 642
990
+ },
991
+ {
992
+ "caption": "perform strutting dance with steps, turns, arm swings; style: rhythmic upbeat",
993
+ "dataset": "100style_original_filtered_test",
994
+ "dataset_index": 2813,
995
+ "generated_matching_distance": 1.3606981039047241,
996
+ "generated_reference_distance_gap": 0.0619969367980957,
997
+ "generated_text_rank": 13,
998
+ "reference_matching_distance": 1.2987011671066284,
999
+ "reference_text_rank": 11,
1000
+ "sample_index": 68
1001
+ },
1002
+ {
1003
+ "caption": "Runners sprint back and forth, fully focused.",
1004
+ "dataset": "fitness_original_filtered_test",
1005
+ "dataset_index": 28,
1006
+ "generated_matching_distance": 1.2401070594787598,
1007
+ "generated_reference_distance_gap": -0.05513405799865723,
1008
+ "generated_text_rank": 1,
1009
+ "reference_matching_distance": 1.295241117477417,
1010
+ "reference_text_rank": 11,
1011
+ "sample_index": 278
1012
+ },
1013
+ {
1014
+ "caption": "stand still facing backward; style: static",
1015
+ "dataset": "amass_original_filtered_test",
1016
+ "dataset_index": 7943,
1017
+ "generated_matching_distance": 1.2716737985610962,
1018
+ "generated_reference_distance_gap": -0.014789938926696777,
1019
+ "generated_text_rank": 8,
1020
+ "reference_matching_distance": 1.286463737487793,
1021
+ "reference_text_rank": 10,
1022
+ "sample_index": 153
1023
+ },
1024
+ {
1025
+ "caption": "stand still, bend forward holding pose; style: static",
1026
+ "dataset": "lafan1_original_filtered_test",
1027
+ "dataset_index": 336,
1028
+ "generated_matching_distance": 1.2387176752090454,
1029
+ "generated_reference_distance_gap": -0.030677199363708496,
1030
+ "generated_text_rank": 2,
1031
+ "reference_matching_distance": 1.269394874572754,
1032
+ "reference_text_rank": 10,
1033
+ "sample_index": 588
1034
+ },
1035
+ {
1036
+ "caption": "a person is standing with their right and left arms in front of them",
1037
+ "dataset": "humanml_original_filtered_test",
1038
+ "dataset_index": 910,
1039
+ "generated_matching_distance": 1.308367133140564,
1040
+ "generated_reference_distance_gap": 0.004720449447631836,
1041
+ "generated_text_rank": 4,
1042
+ "reference_matching_distance": 1.3036466836929321,
1043
+ "reference_text_rank": 9,
1044
+ "sample_index": 352
1045
+ },
1046
+ {
1047
+ "caption": "execute dance with arm swings, steps, spins, and turns; style: rhythmic",
1048
+ "dataset": "100style_original_filtered_test",
1049
+ "dataset_index": 1259,
1050
+ "generated_matching_distance": 1.2143481969833374,
1051
+ "generated_reference_distance_gap": -0.05173027515411377,
1052
+ "generated_text_rank": 2,
1053
+ "reference_matching_distance": 1.2660784721374512,
1054
+ "reference_text_rank": 9,
1055
+ "sample_index": 24
1056
+ },
1057
+ {
1058
+ "caption": "A man practices martial arts, starting with a static stance before executing swift sword movements.",
1059
+ "dataset": "motiongv_original_filtered_test",
1060
+ "dataset_index": 119450,
1061
+ "generated_matching_distance": 1.2796438932418823,
1062
+ "generated_reference_distance_gap": 0.04078948497772217,
1063
+ "generated_text_rank": 5,
1064
+ "reference_matching_distance": 1.2388544082641602,
1065
+ "reference_text_rank": 9,
1066
+ "sample_index": 679
1067
+ },
1068
+ {
1069
+ "caption": "a seated person moves arms and legs in running-in-place motions; style: rhythmic",
1070
+ "dataset": "idea400_original_filtered_test",
1071
+ "dataset_index": 2093,
1072
+ "generated_matching_distance": 1.4057533740997314,
1073
+ "generated_reference_distance_gap": 0.04399609565734863,
1074
+ "generated_text_rank": 12,
1075
+ "reference_matching_distance": 1.3617572784423828,
1076
+ "reference_text_rank": 8,
1077
+ "sample_index": 486
1078
+ },
1079
+ {
1080
+ "caption": "swing and rotate arms in sequential, circular motions; style: dynamic",
1081
+ "dataset": "omomo_original_filtered_test",
1082
+ "dataset_index": 1246,
1083
+ "generated_matching_distance": 1.171722412109375,
1084
+ "generated_reference_distance_gap": -0.15125954151153564,
1085
+ "generated_text_rank": 4,
1086
+ "reference_matching_distance": 1.3229819536209106,
1087
+ "reference_text_rank": 8,
1088
+ "sample_index": 811
1089
+ },
1090
+ {
1091
+ "caption": "a person leaning ove to fix his pant leg with the right hand and then readjusts his pant leg again with the left hand.",
1092
+ "dataset": "humanml_original_filtered_test",
1093
+ "dataset_index": 4597,
1094
+ "generated_matching_distance": 1.2802313566207886,
1095
+ "generated_reference_distance_gap": -0.02388620376586914,
1096
+ "generated_text_rank": 10,
1097
+ "reference_matching_distance": 1.3041175603866577,
1098
+ "reference_text_rank": 8,
1099
+ "sample_index": 398
1100
+ },
1101
+ {
1102
+ "caption": "walk forward with coordinated arm and leg movements; style: steady",
1103
+ "dataset": "omomo_original_filtered_test",
1104
+ "dataset_index": 2228,
1105
+ "generated_matching_distance": 1.1243443489074707,
1106
+ "generated_reference_distance_gap": -0.17923307418823242,
1107
+ "generated_text_rank": 1,
1108
+ "reference_matching_distance": 1.3035774230957031,
1109
+ "reference_text_rank": 8,
1110
+ "sample_index": 837
1111
+ }
1112
+ ]
1113
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/benchmark.json ADDED
@@ -0,0 +1,297 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "batch_size": 32,
3
+ "cfg_scale": null,
4
+ "ckpt_path": "${OMG_REPO_ROOT}/outputs/g1_transformer_guided_300m_mixed_modal_all_original_rot6d_diffusion_only_8gpu_materialized_scaling100k/checkpoints/sstep=050000.ckpt",
5
+ "dataset_metrics": {
6
+ "100style_original_filtered_test": {
7
+ "embedding": {
8
+ "diversity_generated": 1.1364151731473764,
9
+ "motion_fid": 1.045917709132431,
10
+ "motion_kid": {
11
+ "mean": 0.0019250390869158807,
12
+ "std": 0.0
13
+ },
14
+ "num_samples": 86
15
+ },
16
+ "num_samples": 86,
17
+ "physical": {
18
+ "body_jerk_mean": {
19
+ "max": 306.9395446777344,
20
+ "mean": 91.73821926116943,
21
+ "min": 14.676246643066406,
22
+ "std": 41.61339123942521
23
+ },
24
+ "contact_sliding_speed": {
25
+ "max": 1.9230101108551025,
26
+ "mean": 0.7933437027318706,
27
+ "min": 0.0035282010212540627,
28
+ "std": 0.39022621340146474
29
+ },
30
+ "foot_ground_error": {
31
+ "max": 0.04590573161840439,
32
+ "mean": 0.013192172581871409,
33
+ "min": 0.0021116272546350956,
34
+ "std": 0.009524469604494283
35
+ },
36
+ "num_samples": 86
37
+ }
38
+ },
39
+ "amass_original_filtered_test": {
40
+ "embedding": {
41
+ "diversity_generated": 1.2778150547595533,
42
+ "motion_fid": 0.986471537676253,
43
+ "motion_kid": {
44
+ "mean": 0.0008933093560523631,
45
+ "std": 0.0
46
+ },
47
+ "num_samples": 42
48
+ },
49
+ "num_samples": 42,
50
+ "physical": {
51
+ "body_jerk_mean": {
52
+ "max": 265.1717834472656,
53
+ "mean": 82.27915482293992,
54
+ "min": 19.15143394470215,
55
+ "std": 47.60730059009735
56
+ },
57
+ "contact_sliding_speed": {
58
+ "max": 1.6666642427444458,
59
+ "mean": 0.4989402235396916,
60
+ "min": 0.0,
61
+ "std": 0.44309131773839616
62
+ },
63
+ "foot_ground_error": {
64
+ "max": 0.11975693702697754,
65
+ "mean": 0.027045591058032142,
66
+ "min": 0.007385035511106253,
67
+ "std": 0.02046281314840826
68
+ },
69
+ "num_samples": 42
70
+ }
71
+ }
72
+ },
73
+ "dataset_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/dataset_metrics.json",
74
+ "embedding_frames": 60,
75
+ "embedding_metrics": {
76
+ "distribution_reference": "real_motion",
77
+ "distribution_reference_num_samples": 128,
78
+ "distribution_reference_split": "test",
79
+ "diversity_generated": 1.2510810642567618,
80
+ "embedding_frames": 60,
81
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
82
+ "motion_fid": 0.8342456227706498,
83
+ "motion_key": "body_pos_local",
84
+ "motion_kid": {
85
+ "mean": 0.001273666176498267,
86
+ "std": 0.0
87
+ },
88
+ "multimodality": null,
89
+ "num_samples": 128
90
+ },
91
+ "exp": "300m",
92
+ "generated_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/generated_qpos.npz",
93
+ "num_frames": 120,
94
+ "num_samples": 128,
95
+ "physical_metrics": {
96
+ "body_jerk_mean": {
97
+ "max": 306.9395446777344,
98
+ "mean": 88.63446374237537,
99
+ "min": 14.676246643066406,
100
+ "std": 43.896179623882574
101
+ },
102
+ "contact_height_threshold": 0.12,
103
+ "contact_penetration_tolerance": 0.02,
104
+ "contact_sliding_speed": {
105
+ "max": 1.9230101108551025,
106
+ "mean": 0.6967425611219369,
107
+ "min": 0.0,
108
+ "std": 0.43109100791568306
109
+ },
110
+ "foot_ground_error": {
111
+ "max": 0.11975693702697754,
112
+ "mean": 0.01773782551936165,
113
+ "min": 0.0021116272546350956,
114
+ "std": 0.015513028983369554
115
+ },
116
+ "frames": 120,
117
+ "num_samples": 128
118
+ },
119
+ "real_motion_metrics": {
120
+ "diversity_real": 1.242295583132288,
121
+ "motion_fid_reference_eval": 0.8474380295854327,
122
+ "motion_fid_split_half": 0.8474380295854327,
123
+ "motion_kid_reference_eval": {
124
+ "mean": 0.001341022998739838,
125
+ "std": 0.0
126
+ },
127
+ "motion_kid_split_half": {
128
+ "mean": 0.001341022998739838,
129
+ "std": 0.0
130
+ },
131
+ "multimodality": "-",
132
+ "num_reference_samples": 128,
133
+ "num_samples": 128
134
+ },
135
+ "reference_dataset_indices": [
136
+ 553,
137
+ 1081,
138
+ 1673,
139
+ 3337,
140
+ 937,
141
+ 2193,
142
+ 3256,
143
+ 4747,
144
+ 6281,
145
+ 6984,
146
+ 8306,
147
+ 2223,
148
+ 2681,
149
+ 3509,
150
+ 3726,
151
+ 4461,
152
+ 5520,
153
+ 5825,
154
+ 12589,
155
+ 14776,
156
+ 17849,
157
+ 22880,
158
+ 22962,
159
+ 780,
160
+ 6561,
161
+ 7271,
162
+ 1430,
163
+ 900,
164
+ 1436,
165
+ 2786,
166
+ 3945,
167
+ 5477,
168
+ 9998,
169
+ 11627,
170
+ 12135,
171
+ 13402,
172
+ 14488,
173
+ 15764,
174
+ 17375,
175
+ 17730,
176
+ 21743,
177
+ 25300,
178
+ 25658,
179
+ 26124,
180
+ 26961,
181
+ 28025,
182
+ 28580,
183
+ 29065,
184
+ 30448,
185
+ 30907,
186
+ 31205,
187
+ 34835,
188
+ 34903,
189
+ 40452,
190
+ 42562,
191
+ 46623,
192
+ 47604,
193
+ 51167,
194
+ 51365,
195
+ 52679,
196
+ 55585,
197
+ 55626,
198
+ 56799,
199
+ 58820,
200
+ 59247,
201
+ 61411,
202
+ 62538,
203
+ 64651,
204
+ 64919,
205
+ 65363,
206
+ 66399,
207
+ 66723,
208
+ 69520,
209
+ 71969,
210
+ 73758,
211
+ 75362,
212
+ 75378,
213
+ 77147,
214
+ 78085,
215
+ 80180,
216
+ 80605,
217
+ 80848,
218
+ 84911,
219
+ 84947,
220
+ 85083,
221
+ 85362,
222
+ 87933,
223
+ 88522,
224
+ 91664,
225
+ 91973,
226
+ 93742,
227
+ 94666,
228
+ 94813,
229
+ 95200,
230
+ 96780,
231
+ 96822,
232
+ 103128,
233
+ 103785,
234
+ 103937,
235
+ 104434,
236
+ 108479,
237
+ 109409,
238
+ 110354,
239
+ 112056,
240
+ 113692,
241
+ 114223,
242
+ 117679,
243
+ 119195,
244
+ 120246,
245
+ 120435,
246
+ 121227,
247
+ 122617,
248
+ 123224,
249
+ 123790,
250
+ 86,
251
+ 390,
252
+ 1614,
253
+ 1787,
254
+ 2761,
255
+ 1582,
256
+ 5943,
257
+ 6096,
258
+ 8508,
259
+ 9168,
260
+ 1286,
261
+ 1846,
262
+ 1207,
263
+ 1252
264
+ ],
265
+ "reference_datasets": [
266
+ "100style_original_filtered_test",
267
+ "aioz_gdance_original_audio_test",
268
+ "aistpp_original_audio_test",
269
+ "amass_original_filtered_test",
270
+ "bones_seed_original_filtered_test",
271
+ "finedance_original_audio_test",
272
+ "humanml_original_filtered_test",
273
+ "idea400_original_filtered_test",
274
+ "motiongv_original_filtered_test",
275
+ "motionllama_original_filtered_test",
276
+ "omomo_original_filtered_test",
277
+ "snapmogen_original_filtered_test"
278
+ ],
279
+ "reference_frames": 60,
280
+ "reference_num_samples": 128,
281
+ "reference_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/reference_qpos.npz",
282
+ "reference_samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/reference_samples.jsonl",
283
+ "reference_split": "test",
284
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/sample_metrics.jsonl",
285
+ "sample_rankings_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/sample_rankings.json",
286
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/samples.jsonl",
287
+ "seed": 0,
288
+ "split": "test",
289
+ "tracker_executed": {
290
+ "enabled": false
291
+ },
292
+ "transition_metrics": {
293
+ "chunk_length": 60,
294
+ "enabled": false,
295
+ "reason": "disabled"
296
+ }
297
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/dataset_metrics.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "100style_original_filtered_test": {
3
+ "embedding": {
4
+ "diversity_generated": 1.1364151731473764,
5
+ "motion_fid": 1.045917709132431,
6
+ "motion_kid": {
7
+ "mean": 0.0019250390869158807,
8
+ "std": 0.0
9
+ },
10
+ "num_samples": 86
11
+ },
12
+ "num_samples": 86,
13
+ "physical": {
14
+ "body_jerk_mean": {
15
+ "max": 306.9395446777344,
16
+ "mean": 91.73821926116943,
17
+ "min": 14.676246643066406,
18
+ "std": 41.61339123942521
19
+ },
20
+ "contact_sliding_speed": {
21
+ "max": 1.9230101108551025,
22
+ "mean": 0.7933437027318706,
23
+ "min": 0.0035282010212540627,
24
+ "std": 0.39022621340146474
25
+ },
26
+ "foot_ground_error": {
27
+ "max": 0.04590573161840439,
28
+ "mean": 0.013192172581871409,
29
+ "min": 0.0021116272546350956,
30
+ "std": 0.009524469604494283
31
+ },
32
+ "num_samples": 86
33
+ }
34
+ },
35
+ "amass_original_filtered_test": {
36
+ "embedding": {
37
+ "diversity_generated": 1.2778150547595533,
38
+ "motion_fid": 0.986471537676253,
39
+ "motion_kid": {
40
+ "mean": 0.0008933093560523631,
41
+ "std": 0.0
42
+ },
43
+ "num_samples": 42
44
+ },
45
+ "num_samples": 42,
46
+ "physical": {
47
+ "body_jerk_mean": {
48
+ "max": 265.1717834472656,
49
+ "mean": 82.27915482293992,
50
+ "min": 19.15143394470215,
51
+ "std": 47.60730059009735
52
+ },
53
+ "contact_sliding_speed": {
54
+ "max": 1.6666642427444458,
55
+ "mean": 0.4989402235396916,
56
+ "min": 0.0,
57
+ "std": 0.44309131773839616
58
+ },
59
+ "foot_ground_error": {
60
+ "max": 0.11975693702697754,
61
+ "mean": 0.027045591058032142,
62
+ "min": 0.007385035511106253,
63
+ "std": 0.02046281314840826
64
+ },
65
+ "num_samples": 42
66
+ }
67
+ }
68
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/embedding_metrics.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "distribution_reference": "real_motion",
3
+ "distribution_reference_num_samples": 128,
4
+ "distribution_reference_split": "test",
5
+ "diversity_generated": 1.2510810642567618,
6
+ "embedding_frames": 60,
7
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
8
+ "motion_fid": 0.8342456227706498,
9
+ "motion_key": "body_pos_local",
10
+ "motion_kid": {
11
+ "mean": 0.001273666176498267,
12
+ "std": 0.0
13
+ },
14
+ "multimodality": null,
15
+ "num_samples": 128
16
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/metrics.json ADDED
@@ -0,0 +1,352 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "batch_size": 32,
3
+ "benchmark": {
4
+ "batch_size": 32,
5
+ "cfg_scale": null,
6
+ "ckpt_path": "${OMG_REPO_ROOT}/outputs/g1_transformer_guided_300m_mixed_modal_all_original_rot6d_diffusion_only_8gpu_materialized_scaling100k/checkpoints/sstep=050000.ckpt",
7
+ "dataset_metrics": {
8
+ "100style_original_filtered_test": {
9
+ "embedding": {
10
+ "diversity_generated": 1.1364151731473764,
11
+ "motion_fid": 1.045917709132431,
12
+ "motion_kid": {
13
+ "mean": 0.0019250390869158807,
14
+ "std": 0.0
15
+ },
16
+ "num_samples": 86
17
+ },
18
+ "num_samples": 86,
19
+ "physical": {
20
+ "body_jerk_mean": {
21
+ "max": 306.9395446777344,
22
+ "mean": 91.73821926116943,
23
+ "min": 14.676246643066406,
24
+ "std": 41.61339123942521
25
+ },
26
+ "contact_sliding_speed": {
27
+ "max": 1.9230101108551025,
28
+ "mean": 0.7933437027318706,
29
+ "min": 0.0035282010212540627,
30
+ "std": 0.39022621340146474
31
+ },
32
+ "foot_ground_error": {
33
+ "max": 0.04590573161840439,
34
+ "mean": 0.013192172581871409,
35
+ "min": 0.0021116272546350956,
36
+ "std": 0.009524469604494283
37
+ },
38
+ "num_samples": 86
39
+ }
40
+ },
41
+ "amass_original_filtered_test": {
42
+ "embedding": {
43
+ "diversity_generated": 1.2778150547595533,
44
+ "motion_fid": 0.986471537676253,
45
+ "motion_kid": {
46
+ "mean": 0.0008933093560523631,
47
+ "std": 0.0
48
+ },
49
+ "num_samples": 42
50
+ },
51
+ "num_samples": 42,
52
+ "physical": {
53
+ "body_jerk_mean": {
54
+ "max": 265.1717834472656,
55
+ "mean": 82.27915482293992,
56
+ "min": 19.15143394470215,
57
+ "std": 47.60730059009735
58
+ },
59
+ "contact_sliding_speed": {
60
+ "max": 1.6666642427444458,
61
+ "mean": 0.4989402235396916,
62
+ "min": 0.0,
63
+ "std": 0.44309131773839616
64
+ },
65
+ "foot_ground_error": {
66
+ "max": 0.11975693702697754,
67
+ "mean": 0.027045591058032142,
68
+ "min": 0.007385035511106253,
69
+ "std": 0.02046281314840826
70
+ },
71
+ "num_samples": 42
72
+ }
73
+ }
74
+ },
75
+ "dataset_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/dataset_metrics.json",
76
+ "embedding_frames": 60,
77
+ "embedding_metrics": {
78
+ "distribution_reference": "real_motion",
79
+ "distribution_reference_num_samples": 128,
80
+ "distribution_reference_split": "test",
81
+ "diversity_generated": 1.2510810642567618,
82
+ "embedding_frames": 60,
83
+ "evaluator_checkpoint": "${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt",
84
+ "motion_fid": 0.8342456227706498,
85
+ "motion_key": "body_pos_local",
86
+ "motion_kid": {
87
+ "mean": 0.001273666176498267,
88
+ "std": 0.0
89
+ },
90
+ "multimodality": null,
91
+ "num_samples": 128
92
+ },
93
+ "exp": "300m",
94
+ "generated_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/generated_qpos.npz",
95
+ "num_frames": 120,
96
+ "num_samples": 128,
97
+ "physical_metrics": {
98
+ "body_jerk_mean": {
99
+ "max": 306.9395446777344,
100
+ "mean": 88.63446374237537,
101
+ "min": 14.676246643066406,
102
+ "std": 43.896179623882574
103
+ },
104
+ "contact_height_threshold": 0.12,
105
+ "contact_penetration_tolerance": 0.02,
106
+ "contact_sliding_speed": {
107
+ "max": 1.9230101108551025,
108
+ "mean": 0.6967425611219369,
109
+ "min": 0.0,
110
+ "std": 0.43109100791568306
111
+ },
112
+ "foot_ground_error": {
113
+ "max": 0.11975693702697754,
114
+ "mean": 0.01773782551936165,
115
+ "min": 0.0021116272546350956,
116
+ "std": 0.015513028983369554
117
+ },
118
+ "frames": 120,
119
+ "num_samples": 128
120
+ },
121
+ "real_motion_metrics": {
122
+ "diversity_real": 1.242295583132288,
123
+ "motion_fid_reference_eval": 0.8474380295854327,
124
+ "motion_fid_split_half": 0.8474380295854327,
125
+ "motion_kid_reference_eval": {
126
+ "mean": 0.001341022998739838,
127
+ "std": 0.0
128
+ },
129
+ "motion_kid_split_half": {
130
+ "mean": 0.001341022998739838,
131
+ "std": 0.0
132
+ },
133
+ "multimodality": "-",
134
+ "num_reference_samples": 128,
135
+ "num_samples": 128
136
+ },
137
+ "reference_dataset_indices": [
138
+ 553,
139
+ 1081,
140
+ 1673,
141
+ 3337,
142
+ 937,
143
+ 2193,
144
+ 3256,
145
+ 4747,
146
+ 6281,
147
+ 6984,
148
+ 8306,
149
+ 2223,
150
+ 2681,
151
+ 3509,
152
+ 3726,
153
+ 4461,
154
+ 5520,
155
+ 5825,
156
+ 12589,
157
+ 14776,
158
+ 17849,
159
+ 22880,
160
+ 22962,
161
+ 780,
162
+ 6561,
163
+ 7271,
164
+ 1430,
165
+ 900,
166
+ 1436,
167
+ 2786,
168
+ 3945,
169
+ 5477,
170
+ 9998,
171
+ 11627,
172
+ 12135,
173
+ 13402,
174
+ 14488,
175
+ 15764,
176
+ 17375,
177
+ 17730,
178
+ 21743,
179
+ 25300,
180
+ 25658,
181
+ 26124,
182
+ 26961,
183
+ 28025,
184
+ 28580,
185
+ 29065,
186
+ 30448,
187
+ 30907,
188
+ 31205,
189
+ 34835,
190
+ 34903,
191
+ 40452,
192
+ 42562,
193
+ 46623,
194
+ 47604,
195
+ 51167,
196
+ 51365,
197
+ 52679,
198
+ 55585,
199
+ 55626,
200
+ 56799,
201
+ 58820,
202
+ 59247,
203
+ 61411,
204
+ 62538,
205
+ 64651,
206
+ 64919,
207
+ 65363,
208
+ 66399,
209
+ 66723,
210
+ 69520,
211
+ 71969,
212
+ 73758,
213
+ 75362,
214
+ 75378,
215
+ 77147,
216
+ 78085,
217
+ 80180,
218
+ 80605,
219
+ 80848,
220
+ 84911,
221
+ 84947,
222
+ 85083,
223
+ 85362,
224
+ 87933,
225
+ 88522,
226
+ 91664,
227
+ 91973,
228
+ 93742,
229
+ 94666,
230
+ 94813,
231
+ 95200,
232
+ 96780,
233
+ 96822,
234
+ 103128,
235
+ 103785,
236
+ 103937,
237
+ 104434,
238
+ 108479,
239
+ 109409,
240
+ 110354,
241
+ 112056,
242
+ 113692,
243
+ 114223,
244
+ 117679,
245
+ 119195,
246
+ 120246,
247
+ 120435,
248
+ 121227,
249
+ 122617,
250
+ 123224,
251
+ 123790,
252
+ 86,
253
+ 390,
254
+ 1614,
255
+ 1787,
256
+ 2761,
257
+ 1582,
258
+ 5943,
259
+ 6096,
260
+ 8508,
261
+ 9168,
262
+ 1286,
263
+ 1846,
264
+ 1207,
265
+ 1252
266
+ ],
267
+ "reference_datasets": [
268
+ "100style_original_filtered_test",
269
+ "aioz_gdance_original_audio_test",
270
+ "aistpp_original_audio_test",
271
+ "amass_original_filtered_test",
272
+ "bones_seed_original_filtered_test",
273
+ "finedance_original_audio_test",
274
+ "humanml_original_filtered_test",
275
+ "idea400_original_filtered_test",
276
+ "motiongv_original_filtered_test",
277
+ "motionllama_original_filtered_test",
278
+ "omomo_original_filtered_test",
279
+ "snapmogen_original_filtered_test"
280
+ ],
281
+ "reference_frames": 60,
282
+ "reference_num_samples": 128,
283
+ "reference_qpos": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/reference_qpos.npz",
284
+ "reference_samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/reference_samples.jsonl",
285
+ "reference_split": "test",
286
+ "sample_metrics_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/sample_metrics.jsonl",
287
+ "sample_rankings_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/sample_rankings.json",
288
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/samples.jsonl",
289
+ "seed": 0,
290
+ "split": "test",
291
+ "tracker_executed": {
292
+ "enabled": false
293
+ },
294
+ "transition_metrics": {
295
+ "chunk_length": 60,
296
+ "enabled": false,
297
+ "reason": "disabled"
298
+ }
299
+ },
300
+ "ckpts": [
301
+ "${OMG_REPO_ROOT}/outputs/g1_transformer_guided_300m_mixed_modal_all_original_rot6d_diffusion_only_8gpu_materialized_scaling100k/checkpoints/sstep=050000.ckpt"
302
+ ],
303
+ "datasets": [
304
+ "100style_original_filtered_test",
305
+ "amass_original_filtered_test",
306
+ "bones_seed_original_filtered_test",
307
+ "fitness_original_filtered_test",
308
+ "humanml_original_filtered_test",
309
+ "idea400_original_filtered_test",
310
+ "lafan1_original_filtered_test",
311
+ "motiongv_original_filtered_test",
312
+ "motionllama_original_filtered_test",
313
+ "omomo_original_filtered_test",
314
+ "permo_original_filtered_test",
315
+ "snapmogen_original_filtered_test",
316
+ "aioz_gdance_original_audio_test",
317
+ "aistpp_original_audio_test",
318
+ "choreomaster_original_audio_test",
319
+ "compas3d_original_audio_test",
320
+ "opendance_original_audio_test",
321
+ "finedance_original_audio_test",
322
+ "beat2_original_humanref_test"
323
+ ],
324
+ "exp": "300m",
325
+ "num_reference_motions": 128,
326
+ "num_texts": 128,
327
+ "reference_datasets": [
328
+ "100style_original_filtered_test",
329
+ "amass_original_filtered_test",
330
+ "bones_seed_original_filtered_test",
331
+ "fitness_original_filtered_test",
332
+ "humanml_original_filtered_test",
333
+ "idea400_original_filtered_test",
334
+ "lafan1_original_filtered_test",
335
+ "motiongv_original_filtered_test",
336
+ "motionllama_original_filtered_test",
337
+ "omomo_original_filtered_test",
338
+ "permo_original_filtered_test",
339
+ "snapmogen_original_filtered_test",
340
+ "aioz_gdance_original_audio_test",
341
+ "aistpp_original_audio_test",
342
+ "choreomaster_original_audio_test",
343
+ "compas3d_original_audio_test",
344
+ "opendance_original_audio_test",
345
+ "finedance_original_audio_test",
346
+ "beat2_original_humanref_test"
347
+ ],
348
+ "reference_samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/reference_samples.jsonl",
349
+ "reference_split": "test",
350
+ "samples_path": "${OMG_REPO_ROOT}/outputs/benchmark_results_mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/samples.jsonl",
351
+ "split": "test"
352
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/physical_metrics.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "body_jerk_mean": {
3
+ "max": 306.9395446777344,
4
+ "mean": 88.63446374237537,
5
+ "min": 14.676246643066406,
6
+ "std": 43.896179623882574
7
+ },
8
+ "contact_height_threshold": 0.12,
9
+ "contact_penetration_tolerance": 0.02,
10
+ "contact_sliding_speed": {
11
+ "max": 1.9230101108551025,
12
+ "mean": 0.6967425611219369,
13
+ "min": 0.0,
14
+ "std": 0.43109100791568306
15
+ },
16
+ "foot_ground_error": {
17
+ "max": 0.11975693702697754,
18
+ "mean": 0.01773782551936165,
19
+ "min": 0.0021116272546350956,
20
+ "std": 0.015513028983369554
21
+ },
22
+ "frames": 120,
23
+ "num_samples": 128
24
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/real_motion_metrics.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "diversity_real": 1.242295583132288,
3
+ "motion_fid_reference_eval": 0.8474380295854327,
4
+ "motion_fid_split_half": 0.8474380295854327,
5
+ "motion_kid_reference_eval": {
6
+ "mean": 0.001341022998739838,
7
+ "std": 0.0
8
+ },
9
+ "motion_kid_split_half": {
10
+ "mean": 0.001341022998739838,
11
+ "std": 0.0
12
+ },
13
+ "multimodality": "-",
14
+ "num_reference_samples": 128,
15
+ "num_samples": 128
16
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/reference_samples.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/sample_metrics.jsonl ADDED
@@ -0,0 +1,128 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"body_jerk_mean": 67.00100708007812, "caption": "turn to side, lower arms and bring legs together; style: controlled", "contact_sliding_speed": 0.415940523147583, "dataset": "100style_original_filtered_test", "dataset_index": 9, "foot_ground_error": 0.01960536651313305, "sample_index": 0}
2
+ {"body_jerk_mean": 80.21965789794922, "caption": "rotate to front, extend arms and spread legs; style: rhythmic", "contact_sliding_speed": 0.9012015461921692, "dataset": "100style_original_filtered_test", "dataset_index": 18, "foot_ground_error": 0.006513154599815607, "sample_index": 1}
3
+ {"body_jerk_mean": 33.13686752319336, "caption": "hold side stance and shift leg weight; style: steady", "contact_sliding_speed": 0.26054877042770386, "dataset": "100style_original_filtered_test", "dataset_index": 28, "foot_ground_error": 0.01562940515577793, "sample_index": 2}
4
+ {"body_jerk_mean": 136.60653686523438, "caption": "spin in circles, step while keeping hands on hips; style: rhythmic", "contact_sliding_speed": 1.6190179586410522, "dataset": "100style_original_filtered_test", "dataset_index": 56, "foot_ground_error": 0.012503665871918201, "sample_index": 3}
5
+ {"body_jerk_mean": 78.51034545898438, "caption": "rotate body fully and step to face backward; style: stiff, slow", "contact_sliding_speed": 0.910891592502594, "dataset": "100style_original_filtered_test", "dataset_index": 76, "foot_ground_error": 0.009999195113778114, "sample_index": 4}
6
+ {"body_jerk_mean": 93.79772186279297, "caption": "execute repeated turning steps and arm swings angrily; style: aggressive, rhythmic", "contact_sliding_speed": 1.0094605684280396, "dataset": "100style_original_filtered_test", "dataset_index": 98, "foot_ground_error": 0.01391808781772852, "sample_index": 5}
7
+ {"body_jerk_mean": 136.69598388671875, "caption": "keep arms above head, spin and step rhythmically; style: rhythmic", "contact_sliding_speed": 1.2125393152236938, "dataset": "100style_original_filtered_test", "dataset_index": 116, "foot_ground_error": 0.0148954251781106, "sample_index": 6}
8
+ {"body_jerk_mean": 114.89599609375, "caption": "step sideways and twist torso repeatedly, arms behind; style: energetic", "contact_sliding_speed": 1.1204720735549927, "dataset": "100style_original_filtered_test", "dataset_index": 141, "foot_ground_error": 0.006947496440261602, "sample_index": 7}
9
+ {"body_jerk_mean": 75.70713806152344, "caption": "perform continuous balanced dance with stepping and arm motions; style: fluid, balanced", "contact_sliding_speed": 0.42941194772720337, "dataset": "100style_original_filtered_test", "dataset_index": 256, "foot_ground_error": 0.007408611476421356, "sample_index": 8}
10
+ {"body_jerk_mean": 60.719764709472656, "caption": "perform continuous balanced dance with stepping and arm motions; style: fluid, balanced", "contact_sliding_speed": 0.770052969455719, "dataset": "100style_original_filtered_test", "dataset_index": 259, "foot_ground_error": 0.008452767506241798, "sample_index": 9}
11
+ {"body_jerk_mean": 78.30573272705078, "caption": "perform continuous balanced dance with stepping and arm motions; style: fluid, balanced", "contact_sliding_speed": 0.5296463370323181, "dataset": "100style_original_filtered_test", "dataset_index": 279, "foot_ground_error": 0.007841052487492561, "sample_index": 10}
12
+ {"body_jerk_mean": 75.52599334716797, "caption": "extend arms, sway torso while shifting legs apart/close; style: rhythmic", "contact_sliding_speed": 0.472750723361969, "dataset": "100style_original_filtered_test", "dataset_index": 310, "foot_ground_error": 0.008504906669259071, "sample_index": 11}
13
+ {"body_jerk_mean": 86.89344787597656, "caption": "extend arms, cross at chest, step and rotate body; style: dynamic", "contact_sliding_speed": 1.1506714820861816, "dataset": "100style_original_filtered_test", "dataset_index": 433, "foot_ground_error": 0.007457122206687927, "sample_index": 12}
14
+ {"body_jerk_mean": 117.81627655029297, "caption": "repeat circular step sequence with arm swings and body turns; style: rhythmic", "contact_sliding_speed": 1.1441255807876587, "dataset": "100style_original_filtered_test", "dataset_index": 473, "foot_ground_error": 0.007201767526566982, "sample_index": 13}
15
+ {"body_jerk_mean": 82.41751861572266, "caption": "crouch low, step side to side with wide stance; style: low, controlled", "contact_sliding_speed": 0.756416380405426, "dataset": "100style_original_filtered_test", "dataset_index": 604, "foot_ground_error": 0.026241131126880646, "sample_index": 14}
16
+ {"body_jerk_mean": 89.55406188964844, "caption": "crouch, shift weight and step side to side; style: rhythmic", "contact_sliding_speed": 0.918601393699646, "dataset": "100style_original_filtered_test", "dataset_index": 609, "foot_ground_error": 0.014104056172072887, "sample_index": 15}
17
+ {"body_jerk_mean": 68.56842803955078, "caption": "turn body and raise both arms outward; style: smooth", "contact_sliding_speed": 0.6974331140518188, "dataset": "100style_original_filtered_test", "dataset_index": 896, "foot_ground_error": 0.006967470981180668, "sample_index": 16}
18
+ {"body_jerk_mean": 77.73735809326172, "caption": "turn body and raise both arms outward; style: smooth", "contact_sliding_speed": 0.6238415241241455, "dataset": "100style_original_filtered_test", "dataset_index": 897, "foot_ground_error": 0.008520587347447872, "sample_index": 17}
19
+ {"body_jerk_mean": 45.55931854248047, "caption": "perform stylized fairy steps in multiple directions; style: graceful", "contact_sliding_speed": 0.3919636905193329, "dataset": "100style_original_filtered_test", "dataset_index": 929, "foot_ground_error": 0.005531663540750742, "sample_index": 18}
20
+ {"body_jerk_mean": 71.0517349243164, "caption": "perform stylized fairy steps in multiple directions; style: graceful", "contact_sliding_speed": 0.7517004013061523, "dataset": "100style_original_filtered_test", "dataset_index": 959, "foot_ground_error": 0.00697722751647234, "sample_index": 19}
21
+ {"body_jerk_mean": 144.47654724121094, "caption": "alternate leg flicks while spinning in continuous circles; style: rhythmic dynamic", "contact_sliding_speed": 1.316180944442749, "dataset": "100style_original_filtered_test", "dataset_index": 1042, "foot_ground_error": 0.025958498939871788, "sample_index": 20}
22
+ {"body_jerk_mean": 129.85104370117188, "caption": "alternate leg flicks while spinning in continuous circles; style: rhythmic dynamic", "contact_sliding_speed": 0.9055005311965942, "dataset": "100style_original_filtered_test", "dataset_index": 1060, "foot_ground_error": 0.023182053118944168, "sample_index": 21}
23
+ {"body_jerk_mean": 54.99848556518555, "caption": "step backward, lower arms to sides and stand still; style: calm", "contact_sliding_speed": 0.6026706099510193, "dataset": "100style_original_filtered_test", "dataset_index": 1089, "foot_ground_error": 0.007815581746399403, "sample_index": 22}
24
+ {"body_jerk_mean": 68.04824829101562, "caption": "turn left, sweep arms in arc, step side to side rhythmically; style: rhythmic", "contact_sliding_speed": 0.21101772785186768, "dataset": "100style_original_filtered_test", "dataset_index": 1134, "foot_ground_error": 0.004744702950119972, "sample_index": 23}
25
+ {"body_jerk_mean": 96.54129028320312, "caption": "execute dance with arm swings, steps, spins, and turns; style: rhythmic", "contact_sliding_speed": 1.0575028657913208, "dataset": "100style_original_filtered_test", "dataset_index": 1259, "foot_ground_error": 0.009148580022156239, "sample_index": 24}
26
+ {"body_jerk_mean": 57.339786529541016, "caption": "extend arms outward then retract to chest; style: fluid", "contact_sliding_speed": 0.029664693400263786, "dataset": "100style_original_filtered_test", "dataset_index": 1318, "foot_ground_error": 0.02727474458515644, "sample_index": 25}
27
+ {"body_jerk_mean": 74.5478286743164, "caption": "execute alternating karate chops with steps and turns; style: rhythmic martial arts", "contact_sliding_speed": 0.7990440726280212, "dataset": "100style_original_filtered_test", "dataset_index": 1327, "foot_ground_error": 0.018658924847841263, "sample_index": 26}
28
+ {"body_jerk_mean": 104.57305908203125, "caption": "execute alternating karate chops with steps and turns; style: rhythmic martial arts", "contact_sliding_speed": 0.8275144100189209, "dataset": "100style_original_filtered_test", "dataset_index": 1339, "foot_ground_error": 0.036736708134412766, "sample_index": 27}
29
+ {"body_jerk_mean": 89.91545867919922, "caption": "execute alternating karate chops with steps and turns; style: rhythmic martial arts", "contact_sliding_speed": 0.8611111640930176, "dataset": "100style_original_filtered_test", "dataset_index": 1362, "foot_ground_error": 0.04590573161840439, "sample_index": 28}
30
+ {"body_jerk_mean": 80.75251007080078, "caption": "execute alternating karate chops with steps and turns; style: rhythmic martial arts", "contact_sliding_speed": 0.9217734336853027, "dataset": "100style_original_filtered_test", "dataset_index": 1365, "foot_ground_error": 0.04028012603521347, "sample_index": 29}
31
+ {"body_jerk_mean": 72.5741958618164, "caption": "rise from crouch, turn body, extend arms outward; style: transitional", "contact_sliding_speed": 0.8950321674346924, "dataset": "100style_original_filtered_test", "dataset_index": 1403, "foot_ground_error": 0.010329410433769226, "sample_index": 30}
32
+ {"body_jerk_mean": 61.57500457763672, "caption": "lean forward, take steps, rotate body; style: controlled", "contact_sliding_speed": 0.7057388424873352, "dataset": "100style_original_filtered_test", "dataset_index": 1470, "foot_ground_error": 0.004966007079929113, "sample_index": 31}
33
+ {"body_jerk_mean": 68.86087799072266, "caption": "extend arms to sides then retract them; style: fluid", "contact_sliding_speed": 0.0549134835600853, "dataset": "100style_original_filtered_test", "dataset_index": 1485, "foot_ground_error": 0.0043009440414607525, "sample_index": 32}
34
+ {"body_jerk_mean": 58.44475555419922, "caption": "rotate body opposite direction while stepping; style: fluid", "contact_sliding_speed": 0.6112416982650757, "dataset": "100style_original_filtered_test", "dataset_index": 1609, "foot_ground_error": 0.006961645092815161, "sample_index": 33}
35
+ {"body_jerk_mean": 33.36434555053711, "caption": "limp on left leg, twist torso and shift weight side to side; style: limping, controlled", "contact_sliding_speed": 0.040266767144203186, "dataset": "100style_original_filtered_test", "dataset_index": 1673, "foot_ground_error": 0.01964535191655159, "sample_index": 34}
36
+ {"body_jerk_mean": 72.52670288085938, "caption": "limp on right leg, turn and step in various directions; style: controlled limping", "contact_sliding_speed": 0.6696916222572327, "dataset": "100style_original_filtered_test", "dataset_index": 1707, "foot_ground_error": 0.0066049667075276375, "sample_index": 35}
37
+ {"body_jerk_mean": 14.676246643066406, "caption": "stand motionless in upright position; style: static", "contact_sliding_speed": 0.0035282010212540627, "dataset": "100style_original_filtered_test", "dataset_index": 1739, "foot_ground_error": 0.0021532613318413496, "sample_index": 36}
38
+ {"body_jerk_mean": 123.17779541015625, "caption": "march rhythmically with arm swings, steps and turns; style: rhythmic, energetic", "contact_sliding_speed": 1.218740463256836, "dataset": "100style_original_filtered_test", "dataset_index": 1759, "foot_ground_error": 0.007907966151833534, "sample_index": 37}
39
+ {"body_jerk_mean": 184.9546661376953, "caption": "spin repeatedly while jumping and swinging arms in circles; style: rhythmic dynamic", "contact_sliding_speed": 1.3546935319900513, "dataset": "100style_original_filtered_test", "dataset_index": 1832, "foot_ground_error": 0.01629413105547428, "sample_index": 38}
40
+ {"body_jerk_mean": 306.9395446777344, "caption": "spin repeatedly while jumping and swinging arms in circles; style: rhythmic dynamic", "contact_sliding_speed": 1.9230101108551025, "dataset": "100style_original_filtered_test", "dataset_index": 1843, "foot_ground_error": 0.030821723863482475, "sample_index": 39}
41
+ {"body_jerk_mean": 85.38753509521484, "caption": "alternate leg lifts with quick arm movements; style: energetic", "contact_sliding_speed": 0.7288610339164734, "dataset": "100style_original_filtered_test", "dataset_index": 1869, "foot_ground_error": 0.0066989450715482235, "sample_index": 40}
42
+ {"body_jerk_mean": 92.64130401611328, "caption": "turn left, step side to side while bent over; style: controlled", "contact_sliding_speed": 0.6535495519638062, "dataset": "100style_original_filtered_test", "dataset_index": 1879, "foot_ground_error": 0.013152536936104298, "sample_index": 41}
43
+ {"body_jerk_mean": 74.0501480102539, "caption": "turn left, step side to side while bent over; style: controlled", "contact_sliding_speed": 0.5627590417861938, "dataset": "100style_original_filtered_test", "dataset_index": 1882, "foot_ground_error": 0.009073123335838318, "sample_index": 42}
44
+ {"body_jerk_mean": 99.07923889160156, "caption": "spin on heels, lift legs and swing arms; style: dynamic", "contact_sliding_speed": 0.9738694429397583, "dataset": "100style_original_filtered_test", "dataset_index": 1921, "foot_ground_error": 0.00674064178019762, "sample_index": 43}
45
+ {"body_jerk_mean": 91.34962463378906, "caption": "walk in a circle on heels, pivot body; style: smooth", "contact_sliding_speed": 0.8792880773544312, "dataset": "100style_original_filtered_test", "dataset_index": 1936, "foot_ground_error": 0.0070211137644946575, "sample_index": 44}
46
+ {"body_jerk_mean": 70.19570922851562, "caption": "rotate 360 degrees while taking small steps; style: smooth", "contact_sliding_speed": 0.7687453627586365, "dataset": "100style_original_filtered_test", "dataset_index": 2012, "foot_ground_error": 0.026137836277484894, "sample_index": 45}
47
+ {"body_jerk_mean": 60.220672607421875, "caption": "rotate body to face backward and extend arms outward; style: smooth", "contact_sliding_speed": 0.7186462879180908, "dataset": "100style_original_filtered_test", "dataset_index": 2030, "foot_ground_error": 0.008535563945770264, "sample_index": 46}
48
+ {"body_jerk_mean": 87.51476287841797, "caption": "crouched toe-walk toward the right side; style: crouched rhythmic", "contact_sliding_speed": 0.8769672513008118, "dataset": "100style_original_filtered_test", "dataset_index": 2095, "foot_ground_error": 0.009179898537695408, "sample_index": 47}
49
+ {"body_jerk_mean": 57.763057708740234, "caption": "crouched toe-walk backward with body turns; style: crouched controlled", "contact_sliding_speed": 0.6321662664413452, "dataset": "100style_original_filtered_test", "dataset_index": 2147, "foot_ground_error": 0.005855968687683344, "sample_index": 48}
50
+ {"body_jerk_mean": 70.7059097290039, "caption": "rotate 180 degrees to face away from camera; style: smooth", "contact_sliding_speed": 0.7501542568206787, "dataset": "100style_original_filtered_test", "dataset_index": 2185, "foot_ground_error": 0.007140051107853651, "sample_index": 49}
51
+ {"body_jerk_mean": 74.7171401977539, "caption": "shift body side to side with bent leg steps; style: waddling", "contact_sliding_speed": 0.3903679847717285, "dataset": "100style_original_filtered_test", "dataset_index": 2192, "foot_ground_error": 0.004086023196578026, "sample_index": 50}
52
+ {"body_jerk_mean": 86.64373016357422, "caption": "perform full 360-degree spin in place with leg steps; style: smooth", "contact_sliding_speed": 0.9429230690002441, "dataset": "100style_original_filtered_test", "dataset_index": 2241, "foot_ground_error": 0.02442021854221821, "sample_index": 51}
53
+ {"body_jerk_mean": 96.91404724121094, "caption": "extend arms, step legs apart, then hop and shift weight; style: dynamic", "contact_sliding_speed": 1.181304931640625, "dataset": "100style_original_filtered_test", "dataset_index": 2257, "foot_ground_error": 0.015355356968939304, "sample_index": 52}
54
+ {"body_jerk_mean": 24.211515426635742, "caption": "step forward and back, cross arms over torso; style: smooth, deliberate", "contact_sliding_speed": 0.019421422854065895, "dataset": "100style_original_filtered_test", "dataset_index": 2276, "foot_ground_error": 0.0021116272546350956, "sample_index": 53}
55
+ {"body_jerk_mean": 63.72378158569336, "caption": "rotate body 360 degrees and face away from camera; style: smooth", "contact_sliding_speed": 0.8695206046104431, "dataset": "100style_original_filtered_test", "dataset_index": 2323, "foot_ground_error": 0.00952949095517397, "sample_index": 54}
56
+ {"body_jerk_mean": 52.94877243041992, "caption": "turn to face camera then step side to side; style: smooth", "contact_sliding_speed": 0.5201227068901062, "dataset": "100style_original_filtered_test", "dataset_index": 2337, "foot_ground_error": 0.01164337620139122, "sample_index": 55}
57
+ {"body_jerk_mean": 72.14366912841797, "caption": "rotate body to side, step forward, raise right arm; style: smooth", "contact_sliding_speed": 0.8342792391777039, "dataset": "100style_original_filtered_test", "dataset_index": 2368, "foot_ground_error": 0.008277866058051586, "sample_index": 56}
58
+ {"body_jerk_mean": 150.71755981445312, "caption": "quick side steps, rotate body and move hands forward; style: energetic", "contact_sliding_speed": 1.5788081884384155, "dataset": "100style_original_filtered_test", "dataset_index": 2398, "foot_ground_error": 0.012279337272047997, "sample_index": 57}
59
+ {"body_jerk_mean": 48.1912956237793, "caption": "extend both arms horizontally outward; style: smooth", "contact_sliding_speed": 0.03764554113149643, "dataset": "100style_original_filtered_test", "dataset_index": 2410, "foot_ground_error": 0.007212118245661259, "sample_index": 58}
60
+ {"body_jerk_mean": 86.04768371582031, "caption": "lower leg, bend knees, and sway hips rhythmically; style: rhythmic", "contact_sliding_speed": 0.7968769669532776, "dataset": "100style_original_filtered_test", "dataset_index": 2466, "foot_ground_error": 0.00825166329741478, "sample_index": 59}
61
+ {"body_jerk_mean": 78.93553924560547, "caption": "walk in circular path with coordinated arm swings; style: rushed", "contact_sliding_speed": 0.554431676864624, "dataset": "100style_original_filtered_test", "dataset_index": 2521, "foot_ground_error": 0.03080761805176735, "sample_index": 60}
62
+ {"body_jerk_mean": 165.6663055419922, "caption": "execute fast turns and alternating leg kicks; style: rushed", "contact_sliding_speed": 1.5581392049789429, "dataset": "100style_original_filtered_test", "dataset_index": 2526, "foot_ground_error": 0.01668405532836914, "sample_index": 61}
63
+ {"body_jerk_mean": 117.61397552490234, "caption": "step forward/backward, bend torso dynamically; style: rushed", "contact_sliding_speed": 1.193239450454712, "dataset": "100style_original_filtered_test", "dataset_index": 2530, "foot_ground_error": 0.03762182965874672, "sample_index": 62}
64
+ {"body_jerk_mean": 101.71583557128906, "caption": "step forward/backward, bend torso dynamically; style: rushed", "contact_sliding_speed": 0.8143906593322754, "dataset": "100style_original_filtered_test", "dataset_index": 2531, "foot_ground_error": 0.02767130546271801, "sample_index": 63}
65
+ {"body_jerk_mean": 153.45587158203125, "caption": "shuffle side to side, cross and uncross arms quickly; style: rushed", "contact_sliding_speed": 0.9632917046546936, "dataset": "100style_original_filtered_test", "dataset_index": 2536, "foot_ground_error": 0.019954277202486992, "sample_index": 64}
66
+ {"body_jerk_mean": 85.50741577148438, "caption": "rotate body 180°, extend arms outward then retract; style: smooth", "contact_sliding_speed": 0.7513167858123779, "dataset": "100style_original_filtered_test", "dataset_index": 2653, "foot_ground_error": 0.015068062581121922, "sample_index": 65}
67
+ {"body_jerk_mean": 118.82408905029297, "caption": "pivot on foot, turn body while stepping; style: rhythmic", "contact_sliding_speed": 0.89241623878479, "dataset": "100style_original_filtered_test", "dataset_index": 2667, "foot_ground_error": 0.014605364762246609, "sample_index": 66}
68
+ {"body_jerk_mean": 101.91654968261719, "caption": "perform strutting dance with steps, turns, arm swings; style: rhythmic upbeat", "contact_sliding_speed": 1.1932722330093384, "dataset": "100style_original_filtered_test", "dataset_index": 2805, "foot_ground_error": 0.008564887568354607, "sample_index": 67}
69
+ {"body_jerk_mean": 100.37003326416016, "caption": "perform strutting dance with steps, turns, arm swings; style: rhythmic upbeat", "contact_sliding_speed": 1.2356593608856201, "dataset": "100style_original_filtered_test", "dataset_index": 2813, "foot_ground_error": 0.009295584633946419, "sample_index": 68}
70
+ {"body_jerk_mean": 116.5025634765625, "caption": "perform strutting dance with steps, turns, arm swings; style: rhythmic upbeat", "contact_sliding_speed": 1.0851960182189941, "dataset": "100style_original_filtered_test", "dataset_index": 2824, "foot_ground_error": 0.01494005136191845, "sample_index": 69}
71
+ {"body_jerk_mean": 92.89290618896484, "caption": "extend arms and perform breaststroke leg kicks; style: coordinated", "contact_sliding_speed": 0.15362010896205902, "dataset": "100style_original_filtered_test", "dataset_index": 2926, "foot_ground_error": 0.008140143007040024, "sample_index": 70}
72
+ {"body_jerk_mean": 110.07861328125, "caption": "alternate freestyle arm strokes with torso turns; style: rhythmic", "contact_sliding_speed": 0.1782347410917282, "dataset": "100style_original_filtered_test", "dataset_index": 2938, "foot_ground_error": 0.002871684730052948, "sample_index": 71}
73
+ {"body_jerk_mean": 82.43488311767578, "caption": "do backstroke arm circling and body rotations; style: fluid", "contact_sliding_speed": 0.7406783103942871, "dataset": "100style_original_filtered_test", "dataset_index": 2941, "foot_ground_error": 0.010601108893752098, "sample_index": 72}
74
+ {"body_jerk_mean": 217.37188720703125, "caption": "spin rapidly, step, and swing arms widely; style: energetic", "contact_sliding_speed": 1.3226898908615112, "dataset": "100style_original_filtered_test", "dataset_index": 2966, "foot_ground_error": 0.022427357733249664, "sample_index": 73}
75
+ {"body_jerk_mean": 90.84111785888672, "caption": "step in place, swing arms in coordinated motions; style: rhythmic", "contact_sliding_speed": 1.155707597732544, "dataset": "100style_original_filtered_test", "dataset_index": 2971, "foot_ground_error": 0.010211611166596413, "sample_index": 74}
76
+ {"body_jerk_mean": 143.619873046875, "caption": "swing shoulders, rotate torso, and step in circular patterns; style: energetic rhythmic", "contact_sliding_speed": 1.1078784465789795, "dataset": "100style_original_filtered_test", "dataset_index": 2998, "foot_ground_error": 0.00941389799118042, "sample_index": 75}
77
+ {"body_jerk_mean": 64.13981628417969, "caption": "tiptoe sideways with small controlled steps; style: precise", "contact_sliding_speed": 0.1564444750547409, "dataset": "100style_original_filtered_test", "dataset_index": 3067, "foot_ground_error": 0.00683120870962739, "sample_index": 76}
78
+ {"body_jerk_mean": 153.8667755126953, "caption": "jump vertically and spin mid-air; style: acrobatic", "contact_sliding_speed": 0.9542739391326904, "dataset": "100style_original_filtered_test", "dataset_index": 3127, "foot_ground_error": 0.040547147393226624, "sample_index": 77}
79
+ {"body_jerk_mean": 57.49701690673828, "caption": "walk in various directions with arm swings and body turns; style: rhythmic", "contact_sliding_speed": 0.6075807809829712, "dataset": "100style_original_filtered_test", "dataset_index": 3150, "foot_ground_error": 0.005206437781453133, "sample_index": 78}
80
+ {"body_jerk_mean": 56.35883712768555, "caption": "walk forward, extend arms, step and turn left; style: steady", "contact_sliding_speed": 0.629000723361969, "dataset": "100style_original_filtered_test", "dataset_index": 3235, "foot_ground_error": 0.004420981742441654, "sample_index": 79}
81
+ {"body_jerk_mean": 119.4578857421875, "caption": "swirl arms alternately, turn body, step in circles; style: rhythmic, dynamic", "contact_sliding_speed": 0.763212263584137, "dataset": "100style_original_filtered_test", "dataset_index": 3285, "foot_ground_error": 0.010813691653311253, "sample_index": 80}
82
+ {"body_jerk_mean": 122.03912353515625, "caption": "execute repeated wide-legged circular turns and weight shifts; style: rhythmic", "contact_sliding_speed": 1.1874494552612305, "dataset": "100style_original_filtered_test", "dataset_index": 3324, "foot_ground_error": 0.009981783106923103, "sample_index": 81}
83
+ {"body_jerk_mean": 111.9440689086914, "caption": "perform hip-wiggling steps, spins, and body rotations; style: rhythmic", "contact_sliding_speed": 1.1672817468643188, "dataset": "100style_original_filtered_test", "dataset_index": 3353, "foot_ground_error": 0.01178346574306488, "sample_index": 82}
84
+ {"body_jerk_mean": 65.81331634521484, "caption": "rotate body, extend arms, step backward slowly; style: smooth", "contact_sliding_speed": 0.3416193425655365, "dataset": "100style_original_filtered_test", "dataset_index": 3429, "foot_ground_error": 0.005860425997525454, "sample_index": 83}
85
+ {"body_jerk_mean": 79.12821197509766, "caption": "shuffle steps, swing arms, and rotate body; style: zombie-like, stiff, repetitive", "contact_sliding_speed": 0.8466920852661133, "dataset": "100style_original_filtered_test", "dataset_index": 3484, "foot_ground_error": 0.005246656481176615, "sample_index": 84}
86
+ {"body_jerk_mean": 67.46790313720703, "caption": "lower arms and walk with regular steps; style: smooth, normal", "contact_sliding_speed": 0.8620387315750122, "dataset": "100style_original_filtered_test", "dataset_index": 3525, "foot_ground_error": 0.007342224940657616, "sample_index": 85}
87
+ {"body_jerk_mean": 77.31114959716797, "caption": "spin 180 degrees to face camera, raise right arm upward, bend torso forward, lift left leg high, bring both hands toward left knee; style: dynamic", "contact_sliding_speed": 0.8114912509918213, "dataset": "amass_original_filtered_test", "dataset_index": 131, "foot_ground_error": 0.020581569522619247, "sample_index": 86}
88
+ {"body_jerk_mean": 89.0555191040039, "caption": "step forward while repeatedly crossing and swinging arms in front of the body; style: controlled", "contact_sliding_speed": 1.0076794624328613, "dataset": "amass_original_filtered_test", "dataset_index": 143, "foot_ground_error": 0.028244679793715477, "sample_index": 87}
89
+ {"body_jerk_mean": 71.5785903930664, "caption": "extend both arms forward with palms open, move them outward, and step backward with left leg; style: controlled", "contact_sliding_speed": 0.38534170389175415, "dataset": "amass_original_filtered_test", "dataset_index": 253, "foot_ground_error": 0.02465108595788479, "sample_index": 88}
90
+ {"body_jerk_mean": 145.8880615234375, "caption": "descend from jump, land on feet, bend knees, then straighten torso to upright stance; style: controlled", "contact_sliding_speed": 0.9146273136138916, "dataset": "amass_original_filtered_test", "dataset_index": 362, "foot_ground_error": 0.020016590133309364, "sample_index": 89}
91
+ {"body_jerk_mean": 124.14634704589844, "caption": "pivot rapidly, raise left arm upward, then bend torso forward and grasp the raised left leg; style: sharp", "contact_sliding_speed": 0.2834872901439667, "dataset": "amass_original_filtered_test", "dataset_index": 592, "foot_ground_error": 0.022714413702487946, "sample_index": 90}
92
+ {"body_jerk_mean": 44.03581237792969, "caption": "straighten body to upright stance, take small forward steps with arms resting at sides; style: controlled", "contact_sliding_speed": 0.022788843140006065, "dataset": "amass_original_filtered_test", "dataset_index": 598, "foot_ground_error": 0.010723928920924664, "sample_index": 91}
93
+ {"body_jerk_mean": 26.323415756225586, "caption": "turn back to original facing direction, adjust arms and legs to return to neutral standing stance; style: controlled", "contact_sliding_speed": 0.03722573444247246, "dataset": "amass_original_filtered_test", "dataset_index": 734, "foot_ground_error": 0.02617591619491577, "sample_index": 92}
94
+ {"body_jerk_mean": 44.25009536743164, "caption": "land on both feet with bent knees to absorb impact; style: controlled", "contact_sliding_speed": 0.27476221323013306, "dataset": "amass_original_filtered_test", "dataset_index": 817, "foot_ground_error": 0.012193862348794937, "sample_index": 93}
95
+ {"body_jerk_mean": 122.5481185913086, "caption": "walk forward with alternating leg movements and gentle arm swings; style: rhythmic, controlled", "contact_sliding_speed": 0.4366573691368103, "dataset": "amass_original_filtered_test", "dataset_index": 1023, "foot_ground_error": 0.010057251900434494, "sample_index": 94}
96
+ {"body_jerk_mean": 23.59238052368164, "caption": "rotate torso back to front, lower arms gradually to neutral position; style: smooth", "contact_sliding_speed": 0.020598119124770164, "dataset": "amass_original_filtered_test", "dataset_index": 1062, "foot_ground_error": 0.039709609001874924, "sample_index": 95}
97
+ {"body_jerk_mean": 157.1649627685547, "caption": "alternate stepping side to side, swing arms in coordination with leg movements; style: rhythmic, controlled", "contact_sliding_speed": 0.5768324732780457, "dataset": "amass_original_filtered_test", "dataset_index": 1127, "foot_ground_error": 0.009778497740626335, "sample_index": 96}
98
+ {"body_jerk_mean": 32.0805778503418, "caption": "step side to side repeatedly, shifting body weight between right and left feet, arms hanging loosely at sides; style: rhythmic, controlled", "contact_sliding_speed": 0.3344438076019287, "dataset": "amass_original_filtered_test", "dataset_index": 1178, "foot_ground_error": 0.02696472778916359, "sample_index": 97}
99
+ {"body_jerk_mean": 166.37948608398438, "caption": "bend knees to initiate jump, leap upward, perform alternating leg lifts and balancing arm movements while airborne; style: dynamic", "contact_sliding_speed": 0.6942203640937805, "dataset": "amass_original_filtered_test", "dataset_index": 1247, "foot_ground_error": 0.04225792735815048, "sample_index": 98}
100
+ {"body_jerk_mean": 19.15143394470215, "caption": "raise right arm upward to touch the head; style: controlled", "contact_sliding_speed": 0.025154614821076393, "dataset": "amass_original_filtered_test", "dataset_index": 1265, "foot_ground_error": 0.04162478446960449, "sample_index": 99}
101
+ {"body_jerk_mean": 87.28064727783203, "caption": "walk forward with alternating leg movements and coordinated arm swings; style: controlled", "contact_sliding_speed": 0.9785009026527405, "dataset": "amass_original_filtered_test", "dataset_index": 1327, "foot_ground_error": 0.007385035511106253, "sample_index": 100}
102
+ {"body_jerk_mean": 54.45087814331055, "caption": "extend right leg forward in a kicking motion, lean torso left, extend arms for balance; style: dynamic", "contact_sliding_speed": 0.3622299134731293, "dataset": "amass_original_filtered_test", "dataset_index": 1330, "foot_ground_error": 0.016103485599160194, "sample_index": 101}
103
+ {"body_jerk_mean": 76.11148834228516, "caption": "rotate torso, step sideways, swing one arm upward and forward to execute a throw, then adjust posture to stable stand; style: dynamic", "contact_sliding_speed": 0.2111799567937851, "dataset": "amass_original_filtered_test", "dataset_index": 1337, "foot_ground_error": 0.026959039270877838, "sample_index": 102}
104
+ {"body_jerk_mean": 76.8113784790039, "caption": "walk forward with natural arm swings; style: rhythmic", "contact_sliding_speed": 0.9181681275367737, "dataset": "amass_original_filtered_test", "dataset_index": 1640, "foot_ground_error": 0.011887611821293831, "sample_index": 103}
105
+ {"body_jerk_mean": 43.2802734375, "caption": "stop walking, raise right arm, rotate torso left, hold pose; style: controlled", "contact_sliding_speed": 0.45787668228149414, "dataset": "amass_original_filtered_test", "dataset_index": 1644, "foot_ground_error": 0.026305045932531357, "sample_index": 104}
106
+ {"body_jerk_mean": 106.59567260742188, "caption": "spin torso while stepping sideways and adjusting arm positions for balance; style: dynamic", "contact_sliding_speed": 0.9263346791267395, "dataset": "amass_original_filtered_test", "dataset_index": 1882, "foot_ground_error": 0.009321858175098896, "sample_index": 105}
107
+ {"body_jerk_mean": 102.06510925292969, "caption": "spin around in a full circle using leg steps and arm movements to maintain balance; style: dynamic", "contact_sliding_speed": 0.9107708930969238, "dataset": "amass_original_filtered_test", "dataset_index": 1963, "foot_ground_error": 0.0371917299926281, "sample_index": 106}
108
+ {"body_jerk_mean": 265.1717834472656, "caption": "jog in place, alternating leg lifts and arm swings rhythmically; style: rhythmic", "contact_sliding_speed": 1.6666642427444458, "dataset": "amass_original_filtered_test", "dataset_index": 2029, "foot_ground_error": 0.01688998192548752, "sample_index": 107}
109
+ {"body_jerk_mean": 126.03910064697266, "caption": "walk forward, alternating left and right leg steps while swinging arms in coordination with leg movements; style: rhythmic, controlled", "contact_sliding_speed": 1.0184115171432495, "dataset": "amass_original_filtered_test", "dataset_index": 2049, "foot_ground_error": 0.00791759230196476, "sample_index": 108}
110
+ {"body_jerk_mean": 84.64654541015625, "caption": "balance on one leg, extend opposite leg forward and sideways repeatedly, shift body weight; style: deliberate", "contact_sliding_speed": 0.34770822525024414, "dataset": "amass_original_filtered_test", "dataset_index": 2075, "foot_ground_error": 0.03526639938354492, "sample_index": 109}
111
+ {"body_jerk_mean": 144.94496154785156, "caption": "spin body in circles while taking steps, swing arms for balance; style: dynamic", "contact_sliding_speed": 1.4155683517456055, "dataset": "amass_original_filtered_test", "dataset_index": 2170, "foot_ground_error": 0.012227682396769524, "sample_index": 110}
112
+ {"body_jerk_mean": 75.32125091552734, "caption": "stand with slight knee bending, gradually extend both arms outward to fully stretched position; style: controlled", "contact_sliding_speed": 0.024492846801877022, "dataset": "amass_original_filtered_test", "dataset_index": 2324, "foot_ground_error": 0.027086075395345688, "sample_index": 111}
113
+ {"body_jerk_mean": 106.40260314941406, "caption": "walk in place with alternating leg steps and coordinated arm swings; style: rhythmic, smooth", "contact_sliding_speed": 1.2911341190338135, "dataset": "amass_original_filtered_test", "dataset_index": 2378, "foot_ground_error": 0.014084363356232643, "sample_index": 112}
114
+ {"body_jerk_mean": 99.69586944580078, "caption": "walk forward on treadmill with alternating leg steps and synchronized arm swings; style: rhythmic, controlled", "contact_sliding_speed": 0.9697054028511047, "dataset": "amass_original_filtered_test", "dataset_index": 2401, "foot_ground_error": 0.02571909874677658, "sample_index": 113}
115
+ {"body_jerk_mean": 29.860332489013672, "caption": "lower arm and adjust body posture in follow-through motion; style: smooth", "contact_sliding_speed": 0.007746955379843712, "dataset": "amass_original_filtered_test", "dataset_index": 2448, "foot_ground_error": 0.028833147138357162, "sample_index": 114}
116
+ {"body_jerk_mean": 71.22827911376953, "caption": "bend knees, twist torso and hips, shift weight between legs while turning body repeatedly; style: rhythmic", "contact_sliding_speed": 0.3478544354438782, "dataset": "amass_original_filtered_test", "dataset_index": 2778, "foot_ground_error": 0.026005331426858902, "sample_index": 115}
117
+ {"body_jerk_mean": 28.95147705078125, "caption": "repeatedly shake hips side to side, keep arms in raised position, bend and straighten knees in sync with hip movement; style: rhythmic", "contact_sliding_speed": 0.01820143312215805, "dataset": "amass_original_filtered_test", "dataset_index": 2919, "foot_ground_error": 0.024182481691241264, "sample_index": 116}
118
+ {"body_jerk_mean": 108.29864501953125, "caption": "execute torso turns and leg lifts with smooth, coordinated arm motions; style: fluid", "contact_sliding_speed": 0.7444813847541809, "dataset": "amass_original_filtered_test", "dataset_index": 3088, "foot_ground_error": 0.015872061252593994, "sample_index": 117}
119
+ {"body_jerk_mean": 112.375732421875, "caption": "perform coordinated dance steps, shift body weight between legs, rotate torso, and sweep arms in flowing arcs while making gentle spins; style: smooth, rhythmic", "contact_sliding_speed": 0.9923105835914612, "dataset": "amass_original_filtered_test", "dataset_index": 3214, "foot_ground_error": 0.05027060583233833, "sample_index": 118}
120
+ {"body_jerk_mean": 56.714111328125, "caption": "step backward while moving arms outward, then return to standing position with arms fully outstretched; style: controlled", "contact_sliding_speed": 0.5053032040596008, "dataset": "amass_original_filtered_test", "dataset_index": 3277, "foot_ground_error": 0.01746981590986252, "sample_index": 119}
121
+ {"body_jerk_mean": 55.64699935913086, "caption": "walk while rotating body clockwise; style: controlled", "contact_sliding_speed": 0.4767299294471741, "dataset": "amass_original_filtered_test", "dataset_index": 3426, "foot_ground_error": 0.08974239230155945, "sample_index": 120}
122
+ {"body_jerk_mean": 44.59029006958008, "caption": "bend forward, place hands on hips, maintain bent posture with slight movements; style: controlled", "contact_sliding_speed": 0.025804849341511726, "dataset": "amass_original_filtered_test", "dataset_index": 3811, "foot_ground_error": 0.017763443291187286, "sample_index": 121}
123
+ {"body_jerk_mean": 43.32829284667969, "caption": "stand with arms extended horizontally, remain motionless; style: static", "contact_sliding_speed": 0.010634006932377815, "dataset": "amass_original_filtered_test", "dataset_index": 3860, "foot_ground_error": 0.020319633185863495, "sample_index": 122}
124
+ {"body_jerk_mean": 51.377418518066406, "caption": "crouch and lean back, perform repetitive arm and upper body gestures mimicking eating; style: rhythmic", "contact_sliding_speed": 0.0, "dataset": "amass_original_filtered_test", "dataset_index": 3959, "foot_ground_error": 0.11975693702697754, "sample_index": 123}
125
+ {"body_jerk_mean": 60.45087432861328, "caption": "raise arm upward, wave hand while shifting body weight, then lower arm back to body side; style: controlled", "contact_sliding_speed": 0.024343455210328102, "dataset": "amass_original_filtered_test", "dataset_index": 4286, "foot_ground_error": 0.039986055344343185, "sample_index": 124}
126
+ {"body_jerk_mean": 50.25828552246094, "caption": "stand with arms extended outward, then lower arms to body sides; style: static", "contact_sliding_speed": 0.01548338495194912, "dataset": "amass_original_filtered_test", "dataset_index": 4436, "foot_ground_error": 0.018764572218060493, "sample_index": 125}
127
+ {"body_jerk_mean": 99.22977447509766, "caption": "spin in place repeatedly, take small steps, move arms to maintain balance during body rotations; style: rhythmic, coordinated", "contact_sliding_speed": 0.44773563742637634, "dataset": "amass_original_filtered_test", "dataset_index": 4490, "foot_ground_error": 0.033191464841365814, "sample_index": 126}
128
+ {"body_jerk_mean": 51.090476989746094, "caption": "stand motionless with arms extended horizontally outward; style: static", "contact_sliding_speed": 0.014803708530962467, "dataset": "amass_original_filtered_test", "dataset_index": 4594, "foot_ground_error": 0.023717038333415985, "sample_index": 127}
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/sample_rankings.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "num_samples": 128
3
+ }
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/samples.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/summary.md ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Generation Benchmark Summary
2
+
3
+ - ckpts: ['${OMG_REPO_ROOT}/outputs/g1_transformer_guided_300m_mixed_modal_all_original_rot6d_diffusion_only_8gpu_materialized_scaling100k/checkpoints/sstep=050000.ckpt']
4
+ - datasets: ['100style_original_filtered_test', 'amass_original_filtered_test', 'bones_seed_original_filtered_test', 'fitness_original_filtered_test', 'humanml_original_filtered_test', 'idea400_original_filtered_test', 'lafan1_original_filtered_test', 'motiongv_original_filtered_test', 'motionllama_original_filtered_test', 'omomo_original_filtered_test', 'permo_original_filtered_test', 'snapmogen_original_filtered_test', 'aioz_gdance_original_audio_test', 'aistpp_original_audio_test', 'choreomaster_original_audio_test', 'compas3d_original_audio_test', 'opendance_original_audio_test', 'finedance_original_audio_test', 'beat2_original_humanref_test']
5
+ - num_texts: 128
6
+ - reference_split: test
7
+ - num_reference_motions: 128
8
+ - batch_size: 32
9
+ - split: test
10
+ - evaluator_checkpoint: ${OMG_REPO_ROOT}/outputs/evaluator/filtered_original_mixed_modalities_all_text_g1_tmr_bodypos_60f/checkpoints/step_004000.pt
11
+ - enable_text_metrics: False
12
+ - t5_3b_model: ${OMG_MODELS_ROOT}/text_encoders/t5-3b-local
13
+ - text_cache_dir: ${OMG_REPO_ROOT}/outputs/evaluator_text_cache/filtered_original_g1_tmr_bodypos_60f
14
+ - text_batch_size: 16
15
+ - metric_definition: FID(seq) and KID(seq) compare sequence-level evaluator embeddings against real motions sampled from --reference-split, which defaults to the training split. The Real motions row compares eval real motions against this reference baseline; model rows compare generated motions against the same reference baseline.
16
+ - text_metrics: R-precision is computed with fixed batch32 retrieval candidate sets. Retrieval batches are stratified by dataset when dataset metadata is available. Matching and MultiModal Dist are paired embedding distances. Text metrics are computed when --enable_text_metrics is provided; --t5_3b_model and --text_cache_dir default to the paths above.
17
+ - multimodality: Computed only when --multimodality_repeats > 1 by sampling each text multiple times and averaging same-text motion embedding distances.
18
+ - transition_metrics: TextOp-style PJ/AUJ on body/link jerk over transition windows centered at autoregressive chunk seams.
19
+
20
+ | ckpt | FID(seq) | KID(seq) | R@1 | R@2 | R@3 | Diversity | Matching | MultiModal Dist | MultiModality | contact_sliding | foot_ground_error | body_jerk | Tracker g-MPJPE | Tracker MPJPE | Tracker E_vel | Tracker E_acc | PJ | AUJ |
21
+ |---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
22
+ | Real motions | 0.8474 | 0.0013 +/- 0.0000 | N/A | N/A | N/A | 1.2423 | N/A | N/A | - | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
23
+ | sstep=050000.ckpt | **0.8342** | **0.0013 +/- 0.0000** | N/A | N/A | N/A | **1.2511** | N/A | N/A | - | **0.6967 +/- 0.4311** | **0.0177 +/- 0.0155** | **88.6345 +/- 43.8962** | N/A | N/A | N/A | N/A | N/A | N/A |
benchmark/results/mixed_modalities_all_v1/text/omnimotion_300m_mixed_modal_all_scaling_s050000/shards_8/shard_000/transition_metrics.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 60,
3
+ "enabled": false,
4
+ "reason": "disabled"
5
+ }