Knowing commited on
Commit
4229dfd
·
verified ·
1 Parent(s): 2d1d0a6

Add files using upload-large-folder tool

Browse files
Files changed (50) hide show
  1. ABLATION_0225_FreqSelect/main.log +128 -0
  2. ABLATION_0225_FreqSelect/peak_vram_memory.json +6 -0
  3. ABLATION_0225_FreqSelect/train_ddp_process_1.log +66 -0
  4. ABLATION_0225_FreqSelect/train_ddp_process_2.log +66 -0
  5. ABLATION_0225_FreqSelect/train_ddp_process_3.log +66 -0
  6. ABLATION_0225_FreqSelect/train_ddp_process_4.log +66 -0
  7. ABLATION_0225_FreqSelect/train_ddp_process_5.log +66 -0
  8. ABLATION_0225_FreqSelect/train_ddp_process_6.log +66 -0
  9. ABLATION_0225_FreqSelect/train_ddp_process_7.log +66 -0
  10. ABLATION_0225_OURS/main.log +128 -0
  11. ABLATION_0225_OURS/peak_vram_memory.json +6 -0
  12. ABLATION_0225_OURS/train_ddp_process_1.log +66 -0
  13. ABLATION_0225_OURS/train_ddp_process_2.log +66 -0
  14. ABLATION_0225_OURS/train_ddp_process_3.log +66 -0
  15. ABLATION_0225_OURS/train_ddp_process_4.log +66 -0
  16. ABLATION_0225_OURS/train_ddp_process_5.log +66 -0
  17. ABLATION_0225_OURS/train_ddp_process_6.log +66 -0
  18. ABLATION_0225_OURS/train_ddp_process_7.log +66 -0
  19. ABLATION_0225_noRefineModule/train_ddp_process_1.log +66 -0
  20. ABLATION_0225_noRefineModule/train_ddp_process_2.log +66 -0
  21. ABLATION_0225_noRefineModule/train_ddp_process_5.log +66 -0
  22. ABLATION_0225_noRefineModule/train_ddp_process_6.log +66 -0
  23. ABLATION_0225_randomSelect/.hydra/config.yaml +185 -0
  24. ABLATION_0225_randomSelect/.hydra/hydra.yaml +165 -0
  25. ABLATION_0225_randomSelect/.hydra/overrides.yaml +4 -0
  26. ABLATION_0225_randomSelect/train_ddp_process_3.log +60 -0
  27. ABLATION_0225_randomSelect/train_ddp_process_7.log +60 -0
  28. ABLATION_0225_randomSelect/wandb/debug-internal.log +6 -0
  29. ABLATION_0225_randomSelect/wandb/debug.log +19 -0
  30. ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/files/output.log +0 -0
  31. ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/files/requirements.txt +172 -0
  32. ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/files/wandb-metadata.json +93 -0
  33. ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug-core.log +7 -0
  34. ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug-internal.log +6 -0
  35. ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug.log +19 -0
  36. ABLATION_0225_targetTrain/main.log +128 -0
  37. ABLATION_0225_targetTrain/peak_vram_memory.json +6 -0
  38. ABLATION_0225_targetTrain/train_ddp_process_1.log +66 -0
  39. ABLATION_0225_targetTrain/train_ddp_process_2.log +66 -0
  40. ABLATION_0225_targetTrain/train_ddp_process_3.log +66 -0
  41. ABLATION_0225_targetTrain/train_ddp_process_4.log +66 -0
  42. ABLATION_0225_targetTrain/train_ddp_process_5.log +66 -0
  43. ABLATION_0225_targetTrain/train_ddp_process_6.log +66 -0
  44. ABLATION_0225_targetTrain/train_ddp_process_7.log +66 -0
  45. ABLATION_0225_targetTrain_SSR/main.log +128 -0
  46. ABLATION_0225_targetTrain_SSR/peak_vram_memory.json +6 -0
  47. ABLATION_0225_targetTrain_SSR/train_ddp_process_1.log +66 -0
  48. ABLATION_0225_targetTrain_SSR/train_ddp_process_3.log +66 -0
  49. ABLATION_0225_targetTrain_SSR/train_ddp_process_5.log +66 -0
  50. ABLATION_0225_targetTrain_SSR/train_ddp_process_6.log +66 -0
ABLATION_0225_FreqSelect/main.log ADDED
@@ -0,0 +1,128 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 22:26:45,255][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 22:26:51,448][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 22:26:51,448][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 22:27:44,325][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:425: The 'val_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=31` in the `DataLoader` to improve performance.
9
+
10
+ [2026-02-24 22:27:44,327][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
11
+ warnings.warn( # warn only once
12
+
13
+ [2026-02-24 22:27:46,874][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
14
+ result[selector] = overlay
15
+
16
+ [2026-02-24 22:27:46,883][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/utilities/data.py:79: Trying to infer the `batch_size` from an ambiguous collection. The batch size we found is 1. To avoid any miscalculations, use `self.log(..., batch_size=batch_size)`.
17
+
18
+ [2026-02-24 22:27:46,884][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
19
+ warnings.warn(
20
+
21
+ [2026-02-24 22:27:46,884][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
22
+ warnings.warn(msg)
23
+
24
+ [2026-02-24 22:27:48,528][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/functional.py:554: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /pytorch/aten/src/ATen/native/TensorShape.cpp:4322.)
25
+ return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]
26
+
27
+ [2026-02-24 22:27:48,812][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/psnr', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
28
+
29
+ [2026-02-24 22:27:48,814][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/lpips', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
30
+
31
+ [2026-02-24 22:27:48,814][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/ssim', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
32
+
33
+ [2026-02-24 22:27:48,814][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/gaussian_num_ratio', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
34
+
35
+ [2026-02-24 22:27:48,814][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('info/global_step', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
36
+
37
+ [2026-02-24 22:27:57,859][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
38
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
39
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
40
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
41
+
42
+ [2026-02-24 22:27:57,942][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
43
+ result[selector] = overlay
44
+
45
+ [2026-02-24 22:29:33,113][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
46
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
47
+
48
+ [2026-02-24 22:40:40,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
49
+ result[selector] = overlay
50
+
51
+ [2026-02-24 22:43:56,273][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
52
+ result[selector] = overlay
53
+
54
+ [2026-02-24 22:53:29,318][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
55
+ result[selector] = overlay
56
+
57
+ [2026-02-24 22:59:47,452][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
58
+ result[selector] = overlay
59
+
60
+ [2026-02-24 23:06:12,751][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
61
+ result[selector] = overlay
62
+
63
+ [2026-02-24 23:15:41,458][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
64
+ result[selector] = overlay
65
+
66
+ [2026-02-24 23:18:57,458][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
67
+ result[selector] = overlay
68
+
69
+ [2026-02-24 23:31:34,760][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
70
+ result[selector] = overlay
71
+
72
+ [2026-02-24 23:31:38,465][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
73
+ result[selector] = overlay
74
+
75
+ [2026-02-24 23:44:21,241][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
76
+ result[selector] = overlay
77
+
78
+ [2026-02-24 23:47:31,382][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
79
+ result[selector] = overlay
80
+
81
+ [2026-02-24 23:57:04,915][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
82
+ result[selector] = overlay
83
+
84
+ [2026-02-25 00:03:37,630][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
85
+ result[selector] = overlay
86
+
87
+ [2026-02-25 00:10:01,185][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
88
+ result[selector] = overlay
89
+
90
+ [2026-02-25 00:19:40,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
91
+ result[selector] = overlay
92
+
93
+ [2026-02-25 00:22:54,741][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
94
+ result[selector] = overlay
95
+
96
+ [2026-02-25 00:35:31,652][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
97
+ result[selector] = overlay
98
+
99
+ [2026-02-25 00:35:35,995][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
100
+ result[selector] = overlay
101
+
102
+ [2026-02-25 00:48:21,267][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
103
+ result[selector] = overlay
104
+
105
+ [2026-02-25 00:51:35,176][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
106
+ result[selector] = overlay
107
+
108
+ [2026-02-25 01:01:09,352][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
109
+ result[selector] = overlay
110
+
111
+ [2026-02-25 01:07:25,490][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
112
+ result[selector] = overlay
113
+
114
+ [2026-02-25 01:13:46,903][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
115
+ result[selector] = overlay
116
+
117
+ [2026-02-25 01:23:13,175][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
118
+ result[selector] = overlay
119
+
120
+ [2026-02-25 01:26:28,369][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
121
+ result[selector] = overlay
122
+
123
+ [2026-02-25 01:39:43,040][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
124
+ result[selector] = overlay
125
+
126
+ [2026-02-25 01:39:46,583][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
127
+ result[selector] = overlay
128
+
ABLATION_0225_FreqSelect/peak_vram_memory.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "peak_memory_allocated_gb": 103.357,
3
+ "peak_memory_reserved_gb": 134.68,
4
+ "total_elapsed_hours": 3.22,
5
+ "mode": "train"
6
+ }
ABLATION_0225_FreqSelect/train_ddp_process_1.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 22:27:01,831][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 22:27:24,994][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 22:27:24,994][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 22:27:44,326][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 22:27:57,862][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 22:27:57,974][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 22:29:33,113][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 22:40:40,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 22:53:29,318][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 23:06:12,750][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 23:18:57,459][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 23:31:38,465][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 23:44:21,241][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 23:57:04,914][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 00:10:01,182][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 00:22:54,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 00:35:35,992][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 00:48:21,267][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 01:01:09,352][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 01:13:46,903][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 01:26:28,369][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 01:39:46,583][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_FreqSelect/train_ddp_process_2.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 22:27:01,794][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 22:27:32,253][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 22:27:32,253][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 22:27:44,326][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 22:27:57,863][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 22:27:57,968][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 22:29:33,114][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 22:40:40,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 22:53:29,318][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 23:06:12,751][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 23:18:57,458][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 23:31:38,465][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 23:44:21,241][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 23:57:04,914][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 00:10:01,182][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 00:22:54,743][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 00:35:35,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 00:48:21,267][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 01:01:09,352][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 01:13:46,903][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 01:26:28,369][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 01:39:46,583][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_FreqSelect/train_ddp_process_3.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 22:27:01,840][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 22:27:31,676][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 22:27:31,676][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 22:27:44,326][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 22:27:57,857][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 22:27:57,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 22:29:33,134][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 22:40:40,462][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 22:53:29,320][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 23:06:12,751][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 23:18:57,458][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 23:31:38,465][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 23:44:21,249][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 23:57:04,914][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 00:10:01,182][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 00:22:54,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 00:35:35,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 00:48:21,267][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 01:01:09,352][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 01:13:46,903][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 01:26:28,369][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 01:39:46,584][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_FreqSelect/train_ddp_process_4.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 22:27:01,837][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 22:27:25,193][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 22:27:25,193][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 22:27:44,326][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 22:27:57,862][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 22:27:57,971][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 22:29:33,140][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 22:40:40,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 22:53:29,318][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 23:06:12,751][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 23:18:57,458][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 23:31:38,465][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 23:44:21,241][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 23:57:04,914][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 00:10:01,182][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 00:22:54,741][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 00:35:35,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 00:48:21,267][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 01:01:09,352][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 01:13:46,903][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 01:26:28,370][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 01:39:46,583][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_FreqSelect/train_ddp_process_5.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 22:27:01,851][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 22:27:32,385][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 22:27:32,385][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 22:27:44,326][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 22:27:57,114][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 22:27:57,978][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 22:29:33,135][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 22:40:40,461][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 22:53:29,318][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 23:06:12,752][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 23:18:57,458][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 23:31:38,466][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 23:44:21,242][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 23:57:04,914][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 00:10:01,182][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 00:22:54,741][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 00:35:35,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 00:48:21,267][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 01:01:09,354][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 01:13:46,903][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 01:26:28,369][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 01:39:46,584][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_FreqSelect/train_ddp_process_6.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 22:27:01,819][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 22:27:32,305][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 22:27:32,305][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 22:27:44,326][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 22:27:57,344][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 22:27:57,964][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 22:29:33,114][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 22:40:40,462][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 22:53:29,318][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 23:06:12,751][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 23:18:57,458][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 23:31:38,465][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 23:44:21,241][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 23:57:04,914][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 00:10:01,182][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 00:22:54,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 00:35:35,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 00:48:21,267][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 01:01:09,352][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 01:13:46,904][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 01:26:28,369][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 01:39:46,584][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_FreqSelect/train_ddp_process_7.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 22:27:01,765][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 22:27:32,214][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 22:27:32,214][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 22:27:44,326][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 22:27:57,352][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 22:27:57,972][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 22:29:33,114][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 22:40:40,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 22:53:29,318][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 23:06:12,751][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 23:18:57,458][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 23:31:38,467][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 23:44:21,241][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 23:57:04,914][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 00:10:01,182][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 00:22:54,741][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 00:35:35,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 00:48:21,268][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 01:01:09,352][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 01:13:46,903][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 01:26:28,369][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 01:39:46,585][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_OURS/main.log ADDED
@@ -0,0 +1,128 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 19:14:18,385][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 19:14:24,430][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 19:14:24,430][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 19:15:14,654][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:425: The 'val_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=31` in the `DataLoader` to improve performance.
9
+
10
+ [2026-02-24 19:15:14,655][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
11
+ warnings.warn( # warn only once
12
+
13
+ [2026-02-24 19:15:17,207][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
14
+ result[selector] = overlay
15
+
16
+ [2026-02-24 19:15:17,216][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/utilities/data.py:79: Trying to infer the `batch_size` from an ambiguous collection. The batch size we found is 1. To avoid any miscalculations, use `self.log(..., batch_size=batch_size)`.
17
+
18
+ [2026-02-24 19:15:17,217][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
19
+ warnings.warn(
20
+
21
+ [2026-02-24 19:15:17,217][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
22
+ warnings.warn(msg)
23
+
24
+ [2026-02-24 19:15:18,865][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/functional.py:554: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /pytorch/aten/src/ATen/native/TensorShape.cpp:4322.)
25
+ return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]
26
+
27
+ [2026-02-24 19:15:19,148][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/psnr', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
28
+
29
+ [2026-02-24 19:15:19,149][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/lpips', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
30
+
31
+ [2026-02-24 19:15:19,150][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/ssim', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
32
+
33
+ [2026-02-24 19:15:19,150][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/gaussian_num_ratio', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
34
+
35
+ [2026-02-24 19:15:19,150][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('info/global_step', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
36
+
37
+ [2026-02-24 19:15:28,655][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
38
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
39
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
40
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
41
+
42
+ [2026-02-24 19:15:28,731][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
43
+ result[selector] = overlay
44
+
45
+ [2026-02-24 19:17:03,331][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
46
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
47
+
48
+ [2026-02-24 19:28:04,865][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
49
+ result[selector] = overlay
50
+
51
+ [2026-02-24 19:31:20,418][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
52
+ result[selector] = overlay
53
+
54
+ [2026-02-24 19:40:53,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
55
+ result[selector] = overlay
56
+
57
+ [2026-02-24 19:47:10,684][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
58
+ result[selector] = overlay
59
+
60
+ [2026-02-24 19:53:35,645][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
61
+ result[selector] = overlay
62
+
63
+ [2026-02-24 20:03:02,783][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
64
+ result[selector] = overlay
65
+
66
+ [2026-02-24 20:06:17,398][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
67
+ result[selector] = overlay
68
+
69
+ [2026-02-24 20:18:52,619][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
70
+ result[selector] = overlay
71
+
72
+ [2026-02-24 20:18:56,264][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
73
+ result[selector] = overlay
74
+
75
+ [2026-02-24 20:31:37,459][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
76
+ result[selector] = overlay
77
+
78
+ [2026-02-24 20:34:47,502][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
79
+ result[selector] = overlay
80
+
81
+ [2026-02-24 20:44:17,905][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
82
+ result[selector] = overlay
83
+
84
+ [2026-02-24 20:50:50,712][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
85
+ result[selector] = overlay
86
+
87
+ [2026-02-24 20:57:11,485][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
88
+ result[selector] = overlay
89
+
90
+ [2026-02-24 21:06:44,763][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
91
+ result[selector] = overlay
92
+
93
+ [2026-02-24 21:09:57,624][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
94
+ result[selector] = overlay
95
+
96
+ [2026-02-24 21:22:33,649][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
97
+ result[selector] = overlay
98
+
99
+ [2026-02-24 21:22:37,986][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
100
+ result[selector] = overlay
101
+
102
+ [2026-02-24 21:35:22,840][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
103
+ result[selector] = overlay
104
+
105
+ [2026-02-24 21:38:36,652][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
106
+ result[selector] = overlay
107
+
108
+ [2026-02-24 21:48:09,908][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
109
+ result[selector] = overlay
110
+
111
+ [2026-02-24 21:54:26,886][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
112
+ result[selector] = overlay
113
+
114
+ [2026-02-24 22:00:48,561][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
115
+ result[selector] = overlay
116
+
117
+ [2026-02-24 22:10:14,366][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
118
+ result[selector] = overlay
119
+
120
+ [2026-02-24 22:13:29,419][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
121
+ result[selector] = overlay
122
+
123
+ [2026-02-24 22:26:21,789][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
124
+ result[selector] = overlay
125
+
126
+ [2026-02-24 22:26:25,372][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
127
+ result[selector] = overlay
128
+
ABLATION_0225_OURS/peak_vram_memory.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "peak_memory_allocated_gb": 103.417,
3
+ "peak_memory_reserved_gb": 137.797,
4
+ "total_elapsed_hours": 3.2,
5
+ "mode": "train"
6
+ }
ABLATION_0225_OURS/train_ddp_process_1.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 19:14:34,504][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 19:15:01,678][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 19:15:01,679][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 19:15:14,655][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 19:15:28,657][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 19:15:28,768][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 19:17:03,331][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 19:28:04,864][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 19:40:53,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 19:53:35,645][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 20:06:17,398][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 20:18:56,264][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 20:31:37,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 20:44:17,904][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-24 20:57:11,483][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-24 21:09:57,624][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-24 21:22:37,985][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-24 21:35:22,839][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-24 21:48:09,908][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-24 22:00:48,561][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-24 22:13:29,419][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-24 22:26:25,372][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_OURS/train_ddp_process_2.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 19:14:34,642][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 19:14:51,445][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 19:14:51,446][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 19:15:14,654][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 19:15:28,659][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 19:15:28,757][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 19:17:03,331][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 19:28:04,864][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 19:40:53,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 19:53:35,645][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 20:06:17,398][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 20:18:56,264][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 20:31:37,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 20:44:17,904][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-24 20:57:11,483][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-24 21:09:57,625][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-24 21:22:37,984][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-24 21:35:22,839][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-24 21:48:09,908][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-24 22:00:48,562][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-24 22:13:29,419][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-24 22:26:25,372][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_OURS/train_ddp_process_3.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 19:14:34,536][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 19:15:01,502][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 19:15:01,502][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 19:15:14,654][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 19:15:28,654][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 19:15:28,768][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 19:17:03,352][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 19:28:04,864][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 19:40:53,016][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 19:53:35,646][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 20:06:17,398][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 20:18:56,264][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 20:31:37,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 20:44:17,904][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-24 20:57:11,483][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-24 21:09:57,624][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-24 21:22:37,985][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-24 21:35:22,839][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-24 21:48:09,908][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-24 22:00:48,563][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-24 22:13:29,419][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-24 22:26:25,372][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_OURS/train_ddp_process_4.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 19:14:34,575][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 19:15:01,647][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 19:15:01,647][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 19:15:14,654][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 19:15:28,656][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 19:15:28,769][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 19:17:03,357][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 19:28:04,864][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 19:40:53,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 19:53:35,646][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 20:06:17,398][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 20:18:56,264][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 20:31:37,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 20:44:17,904][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-24 20:57:11,483][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-24 21:09:57,623][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-24 21:22:37,985][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-24 21:35:22,839][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-24 21:48:09,908][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-24 22:00:48,561][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-24 22:13:29,418][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-24 22:26:25,372][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_OURS/train_ddp_process_5.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 19:14:34,470][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 19:15:00,362][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 19:15:00,362][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 19:15:14,654][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 19:15:27,920][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 19:15:28,766][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 19:17:03,354][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 19:28:04,864][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 19:40:53,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 19:53:35,646][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 20:06:17,398][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 20:18:56,265][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 20:31:37,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 20:44:17,904][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-24 20:57:11,483][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-24 21:09:57,624][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-24 21:22:37,985][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-24 21:35:22,839][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-24 21:48:09,910][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-24 22:00:48,561][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-24 22:13:29,418][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-24 22:26:25,371][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_OURS/train_ddp_process_6.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 19:14:34,825][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 19:15:01,884][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 19:15:01,884][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 19:15:14,654][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 19:15:28,142][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 19:15:28,765][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 19:17:03,331][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 19:28:04,863][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 19:40:53,014][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 19:53:35,646][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 20:06:17,398][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 20:18:56,264][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 20:31:37,460][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 20:44:17,904][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-24 20:57:11,483][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-24 21:09:57,624][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-24 21:22:37,984][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-24 21:35:22,838][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-24 21:48:09,908][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-24 22:00:48,562][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-24 22:13:29,418][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-24 22:26:25,372][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_OURS/train_ddp_process_7.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-24 19:14:34,642][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-24 19:14:51,312][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-24 19:14:51,312][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-24 19:15:14,654][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-24 19:15:28,148][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-24 19:15:28,766][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-24 19:17:03,331][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-24 19:28:04,864][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-24 19:40:53,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-24 19:53:35,646][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-24 20:06:17,399][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-24 20:18:56,266][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-24 20:31:37,461][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-24 20:44:17,904][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-24 20:57:11,483][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-24 21:09:57,624][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-24 21:22:37,984][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-24 21:35:22,839][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-24 21:48:09,908][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-24 22:00:48,561][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-24 22:13:29,419][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-24 22:26:25,373][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_noRefineModule/train_ddp_process_1.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 07:31:50,753][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 07:32:19,967][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 07:32:19,968][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 07:32:30,542][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 07:32:44,869][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 07:32:44,981][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 07:34:17,416][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 07:45:01,534][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 07:57:27,231][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 08:09:48,815][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 08:22:10,768][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 08:34:29,313][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 08:46:50,775][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 08:59:12,244][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 09:11:48,007][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 09:24:15,287][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 09:36:33,848][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 09:48:56,863][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 10:01:25,665][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 10:13:42,512][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 10:26:05,189][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 10:38:43,134][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_noRefineModule/train_ddp_process_2.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 07:31:50,712][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 07:32:19,862][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 07:32:19,863][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 07:32:30,542][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 07:32:44,875][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 07:32:44,981][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 07:34:17,416][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 07:45:01,533][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 07:57:27,232][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 08:09:48,816][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 08:22:10,769][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 08:34:29,314][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 08:46:50,775][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 08:59:12,244][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 09:11:48,007][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 09:24:15,289][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 09:36:33,849][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 09:48:56,863][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 10:01:25,665][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 10:13:42,512][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 10:26:05,189][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 10:38:43,134][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_noRefineModule/train_ddp_process_5.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 07:31:50,721][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 07:32:20,042][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 07:32:20,043][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 07:32:30,542][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 07:32:44,137][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 07:32:44,980][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 07:34:17,445][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 07:45:01,534][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 07:57:27,231][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 08:09:48,816][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 08:22:10,768][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 08:34:29,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 08:46:50,775][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 08:59:12,243][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 09:11:48,007][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 09:24:15,287][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 09:36:33,848][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 09:48:56,863][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 10:01:25,666][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 10:13:42,512][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 10:26:05,189][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 10:38:43,134][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_noRefineModule/train_ddp_process_6.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 07:31:50,845][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 07:32:13,590][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 07:32:13,591][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 07:32:30,542][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 07:32:44,355][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 07:32:44,980][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 07:34:17,417][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 07:45:01,533][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 07:57:27,230][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 08:09:48,815][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 08:22:10,768][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 08:34:29,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 08:46:50,775][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 08:59:12,360][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 09:11:48,007][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 09:24:15,287][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 09:36:33,848][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 09:48:56,863][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 10:01:25,665][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 10:13:42,512][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 10:26:05,189][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 10:38:43,134][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_randomSelect/.hydra/config.yaml ADDED
@@ -0,0 +1,185 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ model:
2
+ encoder:
3
+ name: dcsplat
4
+ input_image_shape:
5
+ - 518
6
+ - 518
7
+ head_mode: pcd
8
+ num_level: 3
9
+ gs_param_dim: 256
10
+ align_corners: false
11
+ use_voxelize: true
12
+ decoder:
13
+ name: splatting_cuda
14
+ background_color:
15
+ - 0.0
16
+ - 0.0
17
+ - 0.0
18
+ make_scale_invariant: false
19
+ density_control:
20
+ name: density_control_module
21
+ mean_dim: 32
22
+ gs_param_dim: 256
23
+ refinement_layer_num: 1
24
+ num_level: 3
25
+ grad_mode: absgrad
26
+ use_mean_features: true
27
+ refinement_type: voxelize
28
+ refinement_hidden_dim: 32
29
+ aggregation_mode: mean
30
+ num_heads: 1
31
+ score_mode: random
32
+ latent_dim: 128
33
+ num_latents: 64
34
+ num_self_attn_per_block: 2
35
+ voxel_size: 0.001
36
+ aux_refine: false
37
+ refine_error: false
38
+ use_refine_module: true
39
+ voxelize_activate: true
40
+ use_depth: false
41
+ render_loss:
42
+ mse:
43
+ weight: 1.0
44
+ lpips:
45
+ weight: 0.05
46
+ apply_after_step: 0
47
+ density_control_loss:
48
+ error_score:
49
+ weight: 0.01
50
+ log_scale: false
51
+ grad_scale: 10000.0
52
+ mode: original
53
+ direct_loss:
54
+ l1:
55
+ weight: 0.8
56
+ ssim:
57
+ weight: 0.2
58
+ wandb:
59
+ project: DCSplat
60
+ entity: scene-representation-group
61
+ name: ABLATION_0225_randomSelect
62
+ mode: online
63
+ tags:
64
+ - re10k
65
+ - 256x256
66
+ mode: train
67
+ data_loader:
68
+ train:
69
+ num_workers: 16
70
+ persistent_workers: true
71
+ batch_size: 16
72
+ seed: 1234
73
+ test:
74
+ num_workers: 4
75
+ persistent_workers: false
76
+ batch_size: 1
77
+ seed: 2345
78
+ val:
79
+ num_workers: 1
80
+ persistent_workers: true
81
+ batch_size: 1
82
+ seed: 3456
83
+ optimizer:
84
+ lr: 0.0002
85
+ warm_up_steps: 25
86
+ backbone_lr_multiplier: 0.1
87
+ backbone_trainable: T+H
88
+ accumulate: 1
89
+ checkpointing:
90
+ load: null
91
+ every_n_train_steps: 1500
92
+ save_top_k: 2
93
+ save_weights_only: false
94
+ train:
95
+ extended_visualization: false
96
+ print_log_every_n_steps: 10
97
+ camera_loss: 10.0
98
+ one_sample_validation: null
99
+ align_corners: false
100
+ intrinsic_scaling: false
101
+ verbose: false
102
+ beta_dist_param:
103
+ - 0.5
104
+ - 4.0
105
+ use_refine_aux: false
106
+ train_target_set: true
107
+ train_gs_num: 1
108
+ ext_scale_detach: false
109
+ cam_scale_mode: sum
110
+ scene_scale_reg_loss: 0.01
111
+ train_aux: true
112
+ vggt_cam_loss: true
113
+ vggt_distil: false
114
+ context_view_train: false
115
+ test:
116
+ output_path: test/ablation/re10k
117
+ align_pose: false
118
+ pose_align_steps: 100
119
+ rot_opt_lr: 0.005
120
+ trans_opt_lr: 0.005
121
+ compute_scores: true
122
+ save_image: false
123
+ save_video: false
124
+ save_active_mask_image: false
125
+ save_error_score_image: false
126
+ save_compare: false
127
+ pred_intrinsic: false
128
+ error_threshold: 0.4
129
+ error_threshold_list:
130
+ - 0.2
131
+ - 0.4
132
+ - 0.6
133
+ - 0.8
134
+ - 1.0
135
+ threshold_mode: ratio
136
+ nvs_view_N_list:
137
+ - 3
138
+ - 6
139
+ - 16
140
+ - 32
141
+ - 64
142
+ seed: 111123
143
+ trainer:
144
+ max_steps: 3001
145
+ val_check_interval: 250
146
+ gradient_clip_val: 0.5
147
+ num_nodes: 1
148
+ dataset:
149
+ re10k:
150
+ make_baseline_1: true
151
+ relative_pose: true
152
+ augment: true
153
+ background_color:
154
+ - 0.0
155
+ - 0.0
156
+ - 0.0
157
+ overfit_to_scene: null
158
+ skip_bad_shape: true
159
+ view_sampler:
160
+ name: bounded
161
+ num_target_views: 4
162
+ num_context_views: 2
163
+ min_distance_between_context_views: 45
164
+ max_distance_between_context_views: 90
165
+ min_distance_to_context_views: 0
166
+ warm_up_steps: 1000
167
+ initial_min_distance_between_context_views: 25
168
+ initial_max_distance_between_context_views: 25
169
+ same_target_gap: false
170
+ num_target_set: 3
171
+ name: re10k
172
+ roots:
173
+ - datasets/re10k
174
+ input_image_shape:
175
+ - 256
176
+ - 256
177
+ original_image_shape:
178
+ - 360
179
+ - 640
180
+ cameras_are_circular: false
181
+ baseline_min: 0.001
182
+ baseline_max: 10000000000.0
183
+ max_fov: 100.0
184
+ dynamic_context_views: true
185
+ max_context_views_per_gpu: 24
ABLATION_0225_randomSelect/.hydra/hydra.yaml ADDED
@@ -0,0 +1,165 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ hydra:
2
+ run:
3
+ dir: outputs/ablation/re10k/${wandb.name}
4
+ sweep:
5
+ dir: multirun/${now:%Y-%m-%d}/${now:%H-%M-%S}
6
+ subdir: ${hydra.job.num}
7
+ launcher:
8
+ _target_: hydra._internal.core_plugins.basic_launcher.BasicLauncher
9
+ sweeper:
10
+ _target_: hydra._internal.core_plugins.basic_sweeper.BasicSweeper
11
+ max_batch_size: null
12
+ params: null
13
+ help:
14
+ app_name: ${hydra.job.name}
15
+ header: '${hydra.help.app_name} is powered by Hydra.
16
+
17
+ '
18
+ footer: 'Powered by Hydra (https://hydra.cc)
19
+
20
+ Use --hydra-help to view Hydra specific help
21
+
22
+ '
23
+ template: '${hydra.help.header}
24
+
25
+ == Configuration groups ==
26
+
27
+ Compose your configuration from those groups (group=option)
28
+
29
+
30
+ $APP_CONFIG_GROUPS
31
+
32
+
33
+ == Config ==
34
+
35
+ Override anything in the config (foo.bar=value)
36
+
37
+
38
+ $CONFIG
39
+
40
+
41
+ ${hydra.help.footer}
42
+
43
+ '
44
+ hydra_help:
45
+ template: 'Hydra (${hydra.runtime.version})
46
+
47
+ See https://hydra.cc for more info.
48
+
49
+
50
+ == Flags ==
51
+
52
+ $FLAGS_HELP
53
+
54
+
55
+ == Configuration groups ==
56
+
57
+ Compose your configuration from those groups (For example, append hydra/job_logging=disabled
58
+ to command line)
59
+
60
+
61
+ $HYDRA_CONFIG_GROUPS
62
+
63
+
64
+ Use ''--cfg hydra'' to Show the Hydra config.
65
+
66
+ '
67
+ hydra_help: ???
68
+ hydra_logging:
69
+ version: 1
70
+ formatters:
71
+ simple:
72
+ format: '[%(asctime)s][HYDRA] %(message)s'
73
+ handlers:
74
+ console:
75
+ class: logging.StreamHandler
76
+ formatter: simple
77
+ stream: ext://sys.stdout
78
+ root:
79
+ level: INFO
80
+ handlers:
81
+ - console
82
+ loggers:
83
+ logging_example:
84
+ level: DEBUG
85
+ disable_existing_loggers: false
86
+ job_logging:
87
+ version: 1
88
+ formatters:
89
+ simple:
90
+ format: '[%(asctime)s][%(name)s][%(levelname)s] - %(message)s'
91
+ handlers:
92
+ console:
93
+ class: logging.StreamHandler
94
+ formatter: simple
95
+ stream: ext://sys.stdout
96
+ file:
97
+ class: logging.FileHandler
98
+ formatter: simple
99
+ filename: ${hydra.runtime.output_dir}/${hydra.job.name}.log
100
+ root:
101
+ level: INFO
102
+ handlers:
103
+ - console
104
+ - file
105
+ disable_existing_loggers: false
106
+ env: {}
107
+ mode: RUN
108
+ searchpath: []
109
+ callbacks: {}
110
+ output_subdir: .hydra
111
+ overrides:
112
+ hydra:
113
+ - hydra.mode=RUN
114
+ task:
115
+ - +experiment=re10k_ablation_24v
116
+ - wandb.mode=online
117
+ - wandb.name=ABLATION_0225_randomSelect
118
+ - model.density_control.score_mode=random
119
+ job:
120
+ name: main
121
+ chdir: null
122
+ override_dirname: +experiment=re10k_ablation_24v,model.density_control.score_mode=random,wandb.mode=online,wandb.name=ABLATION_0225_randomSelect
123
+ id: ???
124
+ num: ???
125
+ config_name: main
126
+ env_set: {}
127
+ env_copy: []
128
+ config:
129
+ override_dirname:
130
+ kv_sep: '='
131
+ item_sep: ','
132
+ exclude_keys: []
133
+ runtime:
134
+ version: 1.3.2
135
+ version_base: '1.3'
136
+ cwd: /workspace/code/CVPR2026
137
+ config_sources:
138
+ - path: hydra.conf
139
+ schema: pkg
140
+ provider: hydra
141
+ - path: /workspace/code/CVPR2026/config
142
+ schema: file
143
+ provider: main
144
+ - path: ''
145
+ schema: structured
146
+ provider: schema
147
+ output_dir: /workspace/code/CVPR2026/outputs/ablation/re10k/ABLATION_0225_randomSelect
148
+ choices:
149
+ experiment: re10k_ablation_24v
150
+ dataset@dataset.re10k: re10k
151
+ dataset/view_sampler_dataset_specific_config@dataset.re10k.view_sampler: bounded_re10k
152
+ dataset/view_sampler@dataset.re10k.view_sampler: bounded
153
+ model/density_control: density_control_module
154
+ model/decoder: splatting_cuda
155
+ model/encoder: dcsplat
156
+ hydra/env: default
157
+ hydra/callbacks: null
158
+ hydra/job_logging: default
159
+ hydra/hydra_logging: default
160
+ hydra/hydra_help: default
161
+ hydra/help: default
162
+ hydra/sweeper: basic
163
+ hydra/launcher: basic
164
+ hydra/output: default
165
+ verbose: false
ABLATION_0225_randomSelect/.hydra/overrides.yaml ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ - +experiment=re10k_ablation_24v
2
+ - wandb.mode=online
3
+ - wandb.name=ABLATION_0225_randomSelect
4
+ - model.density_control.score_mode=random
ABLATION_0225_randomSelect/train_ddp_process_3.log ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 10:39:19,835][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 10:39:48,225][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 10:39:48,225][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 10:39:59,701][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 10:40:13,355][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 10:40:13,462][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 10:41:49,301][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 10:52:55,865][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 11:05:44,779][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 11:18:29,815][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 11:31:13,250][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 11:43:54,528][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 11:56:39,756][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 12:09:23,940][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 12:22:24,117][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 12:35:14,407][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 12:47:56,084][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 13:00:42,869][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 13:13:32,568][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 13:26:12,526][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
ABLATION_0225_randomSelect/train_ddp_process_7.log ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 10:39:19,892][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 10:39:48,493][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 10:39:48,493][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 10:39:59,701][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 10:40:12,849][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [256, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [256, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 10:40:13,459][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 10:41:49,277][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 10:52:55,864][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 11:05:44,777][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 11:18:29,814][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 11:31:13,251][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 11:43:54,528][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 11:56:39,755][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 12:09:23,940][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 12:22:24,117][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 12:35:14,407][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 12:47:56,084][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 13:00:42,870][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 13:13:32,568][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 13:26:12,524][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
ABLATION_0225_randomSelect/wandb/debug-internal.log ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {"time":"2026-02-25T10:39:55.743497828Z","level":"INFO","msg":"stream: starting","core version":"0.25.0"}
2
+ {"time":"2026-02-25T10:39:56.464084002Z","level":"INFO","msg":"stream: created new stream","id":"n76tgbfj"}
3
+ {"time":"2026-02-25T10:39:56.464470748Z","level":"INFO","msg":"handler: started","stream_id":"n76tgbfj"}
4
+ {"time":"2026-02-25T10:39:56.464733133Z","level":"INFO","msg":"stream: started","id":"n76tgbfj"}
5
+ {"time":"2026-02-25T10:39:56.464769993Z","level":"INFO","msg":"sender: started","stream_id":"n76tgbfj"}
6
+ {"time":"2026-02-25T10:39:56.464783063Z","level":"INFO","msg":"writer: started","stream_id":"n76tgbfj"}
ABLATION_0225_randomSelect/wandb/debug.log ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2026-02-25 10:39:55,444 INFO MainThread:145562 [wandb_setup.py:_flush():81] Current SDK version is 0.25.0
2
+ 2026-02-25 10:39:55,444 INFO MainThread:145562 [wandb_setup.py:_flush():81] Configure stats pid to 145562
3
+ 2026-02-25 10:39:55,444 INFO MainThread:145562 [wandb_setup.py:_flush():81] Loading settings from environment variables
4
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:setup_run_log_directory():717] Logging user logs to /workspace/code/CVPR2026/outputs/ablation/re10k/ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug.log
5
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:setup_run_log_directory():718] Logging internal logs to /workspace/code/CVPR2026/outputs/ablation/re10k/ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug-internal.log
6
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:init():844] calling init triggers
7
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:init():849] wandb.init called with sweep_config: {}
8
+ config: {'model': {'encoder': {'name': 'dcsplat', 'input_image_shape': [518, 518], 'head_mode': 'pcd', 'num_level': 3, 'gs_param_dim': 256, 'align_corners': False, 'use_voxelize': True}, 'decoder': {'name': 'splatting_cuda', 'background_color': [0.0, 0.0, 0.0], 'make_scale_invariant': False}, 'density_control': {'name': 'density_control_module', 'mean_dim': 32, 'gs_param_dim': 256, 'refinement_layer_num': 1, 'num_level': 3, 'grad_mode': 'absgrad', 'use_mean_features': True, 'refinement_type': 'voxelize', 'refinement_hidden_dim': 32, 'aggregation_mode': 'mean', 'num_heads': 1, 'score_mode': 'random', 'latent_dim': 128, 'num_latents': 64, 'num_self_attn_per_block': 2, 'voxel_size': 0.001, 'aux_refine': False, 'refine_error': False, 'use_refine_module': True, 'voxelize_activate': True, 'use_depth': False}}, 'render_loss': {'mse': {'weight': 1.0}, 'lpips': {'weight': 0.05, 'apply_after_step': 0}}, 'density_control_loss': {'error_score': {'weight': 0.01, 'log_scale': False, 'grad_scale': 10000.0, 'mode': 'original'}}, 'direct_loss': {'l1': {'weight': 0.8}, 'ssim': {'weight': 0.2}}, 'wandb': {'project': 'DCSplat', 'entity': 'scene-representation-group', 'name': 'ABLATION_0225_randomSelect', 'mode': 'online', 'tags': ['re10k', '256x256']}, 'mode': 'train', 'data_loader': {'train': {'num_workers': 16, 'persistent_workers': True, 'batch_size': 16, 'seed': 1234}, 'test': {'num_workers': 4, 'persistent_workers': False, 'batch_size': 1, 'seed': 2345}, 'val': {'num_workers': 1, 'persistent_workers': True, 'batch_size': 1, 'seed': 3456}}, 'optimizer': {'lr': 0.0002, 'warm_up_steps': 25, 'backbone_lr_multiplier': 0.1, 'backbone_trainable': 'T+H', 'accumulate': 1}, 'checkpointing': {'load': None, 'every_n_train_steps': 1500, 'save_top_k': 2, 'save_weights_only': False}, 'train': {'extended_visualization': False, 'print_log_every_n_steps': 10, 'camera_loss': 10.0, 'one_sample_validation': None, 'align_corners': False, 'intrinsic_scaling': False, 'verbose': False, 'beta_dist_param': [0.5, 4.0], 'use_refine_aux': False, 'train_target_set': True, 'train_gs_num': 1, 'ext_scale_detach': False, 'cam_scale_mode': 'sum', 'scene_scale_reg_loss': 0.01, 'train_aux': True, 'vggt_cam_loss': True, 'vggt_distil': False, 'context_view_train': False}, 'test': {'output_path': 'test/ablation/re10k', 'align_pose': False, 'pose_align_steps': 100, 'rot_opt_lr': 0.005, 'trans_opt_lr': 0.005, 'compute_scores': True, 'save_image': False, 'save_video': False, 'save_active_mask_image': False, 'save_error_score_image': False, 'save_compare': False, 'pred_intrinsic': False, 'error_threshold': 0.4, 'error_threshold_list': [0.2, 0.4, 0.6, 0.8, 1.0], 'threshold_mode': 'ratio', 'nvs_view_N_list': [3, 6, 16, 32, 64]}, 'seed': 111123, 'trainer': {'max_steps': 3001, 'val_check_interval': 250, 'gradient_clip_val': 0.5, 'num_nodes': 1}, 'dataset': {'re10k': {'make_baseline_1': True, 'relative_pose': True, 'augment': True, 'background_color': [0.0, 0.0, 0.0], 'overfit_to_scene': None, 'skip_bad_shape': True, 'view_sampler': {'name': 'bounded', 'num_target_views': 4, 'num_context_views': 2, 'min_distance_between_context_views': 45, 'max_distance_between_context_views': 90, 'min_distance_to_context_views': 0, 'warm_up_steps': 1000, 'initial_min_distance_between_context_views': 25, 'initial_max_distance_between_context_views': 25, 'same_target_gap': False, 'num_target_set': 3}, 'name': 're10k', 'roots': ['datasets/re10k'], 'input_image_shape': [256, 256], 'original_image_shape': [360, 640], 'cameras_are_circular': False, 'baseline_min': 0.001, 'baseline_max': 10000000000.0, 'max_fov': 100.0, 'dynamic_context_views': True, 'max_context_views_per_gpu': 24}}, '_wandb': {}}
9
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:init():892] starting backend
10
+ 2026-02-25 10:39:55,732 INFO MainThread:145562 [wandb_init.py:init():895] sending inform_init request
11
+ 2026-02-25 10:39:55,739 INFO MainThread:145562 [wandb_init.py:init():903] backend started and connected
12
+ 2026-02-25 10:39:55,744 INFO MainThread:145562 [wandb_init.py:init():973] updated telemetry
13
+ 2026-02-25 10:39:55,752 INFO MainThread:145562 [wandb_init.py:init():997] communicating run to backend with 90.0 second timeout
14
+ 2026-02-25 10:39:57,667 INFO MainThread:145562 [wandb_init.py:init():1042] starting run threads in backend
15
+ 2026-02-25 10:39:57,796 INFO MainThread:145562 [wandb_run.py:_console_start():2524] atexit reg
16
+ 2026-02-25 10:39:57,797 INFO MainThread:145562 [wandb_run.py:_redirect():2373] redirect: wrap_raw
17
+ 2026-02-25 10:39:57,797 INFO MainThread:145562 [wandb_run.py:_redirect():2442] Wrapping output streams.
18
+ 2026-02-25 10:39:57,797 INFO MainThread:145562 [wandb_run.py:_redirect():2465] Redirects installed.
19
+ 2026-02-25 10:39:57,800 INFO MainThread:145562 [wandb_init.py:init():1082] run started, returning control to user process
ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/files/output.log ADDED
The diff for this file is too large to render. See raw diff
 
ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/files/requirements.txt ADDED
@@ -0,0 +1,172 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ wheel==0.45.1
2
+ pytz==2025.2
3
+ easydict==1.13
4
+ antlr4-python3-runtime==4.9.3
5
+ wadler_lindig==0.1.7
6
+ urllib3==2.5.0
7
+ tzdata==2025.2
8
+ typing-inspection==0.4.1
9
+ tabulate==0.9.0
10
+ smmap==5.0.2
11
+ kornia_rs==0.1.9
12
+ setuptools==78.1.1
13
+ safetensors==0.5.3
14
+ PyYAML==6.0.2
15
+ PySocks==1.7.1
16
+ pyparsing==3.2.5
17
+ pydantic_core==2.33.2
18
+ pycparser==2.23
19
+ protobuf==6.32.1
20
+ propcache==0.3.2
21
+ proglog==0.1.12
22
+ fsspec==2024.6.1
23
+ platformdirs==4.4.0
24
+ pip==25.2
25
+ pillow==10.4.0
26
+ frozenlist==1.7.0
27
+ packaging==24.2
28
+ opt_einsum==3.4.0
29
+ numpy==1.26.4
30
+ ninja==1.13.0
31
+ fonttools==4.60.0
32
+ networkx==3.4.2
33
+ multidict==6.6.4
34
+ mdurl==0.1.2
35
+ MarkupSafe==3.0.2
36
+ kiwisolver==1.4.9
37
+ imageio-ffmpeg==0.6.0
38
+ idna==3.7
39
+ hf-xet==1.1.10
40
+ gmpy2==2.2.1
41
+ einops==0.8.1
42
+ filelock==3.17.0
43
+ decorator==4.4.2
44
+ dacite==1.9.2
45
+ cycler==0.12.1
46
+ colorama==0.4.6
47
+ click==8.3.0
48
+ nvidia-nvtx-cu12==12.8.90
49
+ charset-normalizer==3.3.2
50
+ certifi==2025.8.3
51
+ beartype==0.19.0
52
+ attrs==25.3.0
53
+ async-timeout==5.0.1
54
+ annotated-types==0.7.0
55
+ aiohappyeyeballs==2.6.1
56
+ yarl==1.20.1
57
+ tifffile==2025.5.10
58
+ sentry-sdk==2.39.0
59
+ scipy==1.15.3
60
+ pydantic==2.11.9
61
+ pandas==2.3.2
62
+ opencv-python==4.11.0.86
63
+ omegaconf==2.3.0
64
+ markdown-it-py==4.0.0
65
+ lightning-utilities==0.14.3
66
+ lazy_loader==0.4
67
+ jaxtyping==0.2.37
68
+ imageio==2.37.0
69
+ gitdb==4.0.12
70
+ contourpy==1.3.2
71
+ colorspacious==1.1.2
72
+ cffi==1.17.1
73
+ aiosignal==1.4.0
74
+ scikit-video==1.1.11
75
+ scikit-image==0.25.2
76
+ rich==14.1.0
77
+ moviepy==1.0.3
78
+ matplotlib==3.10.6
79
+ hydra-core==1.3.2
80
+ nvidia-nccl-cu12==2.27.3
81
+ huggingface-hub==0.35.1
82
+ GitPython==3.1.45
83
+ brotlicffi==1.0.9.2
84
+ aiohttp==3.12.15
85
+ torchmetrics==1.8.2
86
+ opt-einsum-fx==0.1.4
87
+ kornia==0.8.1
88
+ pytorch-lightning==2.5.1
89
+ lpips==0.1.4
90
+ e3nn==0.6.0
91
+ lightning==2.5.1
92
+ nvidia-cusparselt-cu12==0.7.1
93
+ triton==3.4.0
94
+ nvidia-nvjitlink-cu12==12.8.93
95
+ nvidia-curand-cu12==10.3.9.90
96
+ nvidia-cufile-cu12==1.13.1.3
97
+ nvidia-cuda-runtime-cu12==12.8.90
98
+ nvidia-cuda-nvrtc-cu12==12.8.93
99
+ nvidia-cuda-cupti-cu12==12.8.90
100
+ nvidia-cublas-cu12==12.8.4.1
101
+ nvidia-cusparse-cu12==12.5.8.93
102
+ nvidia-cufft-cu12==11.3.3.83
103
+ nvidia-cudnn-cu12==9.10.2.21
104
+ nvidia-cusolver-cu12==11.7.3.90
105
+ torch==2.8.0+cu128
106
+ torchvision==0.23.0+cu128
107
+ torchaudio==2.8.0+cu128
108
+ torch_scatter==2.1.2+pt28cu128
109
+ gsplat==1.5.3
110
+ wandb==0.25.0
111
+ cuda-bindings==13.0.3
112
+ cuda-pathfinder==1.3.3
113
+ Jinja2==3.1.6
114
+ mpmath==1.3.0
115
+ nvidia-cublas==13.1.0.3
116
+ nvidia-cuda-cupti==13.0.85
117
+ nvidia-cuda-nvrtc==13.0.88
118
+ nvidia-cuda-runtime==13.0.96
119
+ nvidia-cudnn-cu13==9.15.1.9
120
+ nvidia-cufft==12.0.0.61
121
+ nvidia-cufile==1.15.1.6
122
+ nvidia-curand==10.4.0.35
123
+ nvidia-cusolver==12.0.4.66
124
+ nvidia-cusparse==12.6.3.3
125
+ nvidia-cusparselt-cu13==0.8.0
126
+ nvidia-nccl-cu13==2.28.9
127
+ nvidia-nvjitlink==13.0.88
128
+ nvidia-nvshmem-cu13==3.4.5
129
+ nvidia-nvtx==13.0.85
130
+ requests==2.32.5
131
+ sentencepiece==0.2.1
132
+ sympy==1.14.0
133
+ torchcodec==0.10.0
134
+ torchdata==0.10.0
135
+ torchtext==0.6.0
136
+ anyio==4.12.0
137
+ asttokens==3.0.1
138
+ comm==0.2.3
139
+ debugpy==1.8.19
140
+ executing==2.2.1
141
+ h11==0.16.0
142
+ httpcore==1.0.9
143
+ httpx==0.28.1
144
+ ipykernel==7.1.0
145
+ ipython==9.8.0
146
+ ipython_pygments_lexers==1.1.1
147
+ ipywidgets==8.1.8
148
+ jedi==0.19.2
149
+ jupyter_client==8.7.0
150
+ jupyter_core==5.9.1
151
+ jupyterlab_widgets==3.0.16
152
+ matplotlib-inline==0.2.1
153
+ nest-asyncio==1.6.0
154
+ parso==0.8.5
155
+ pexpect==4.9.0
156
+ prompt_toolkit==3.0.52
157
+ psutil==7.2.1
158
+ ptyprocess==0.7.0
159
+ pure_eval==0.2.3
160
+ Pygments==2.19.2
161
+ python-dateutil==2.9.0.post0
162
+ pyzmq==27.1.0
163
+ shellingham==1.5.4
164
+ six==1.17.0
165
+ stack-data==0.6.3
166
+ tornado==6.5.4
167
+ tqdm==4.67.1
168
+ traitlets==5.14.3
169
+ typer-slim==0.21.0
170
+ typing_extensions==4.15.0
171
+ wcwidth==0.2.14
172
+ widgetsnbextension==4.0.15
ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/files/wandb-metadata.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "os": "Linux-6.8.0-90-generic-x86_64-with-glibc2.39",
3
+ "python": "CPython 3.12.12",
4
+ "startedAt": "2026-02-25T10:39:55.441642Z",
5
+ "args": [
6
+ "+experiment=re10k_ablation_24v",
7
+ "wandb.mode=online",
8
+ "wandb.name=ABLATION_0225_randomSelect",
9
+ "model.density_control.score_mode=random"
10
+ ],
11
+ "program": "-m src.main",
12
+ "git": {
13
+ "remote": "git@github.com:K-nowing/CVPR2026.git",
14
+ "commit": "2512754c6c27ca5150bf17fbcbdde3f192fd53cc"
15
+ },
16
+ "email": "dna9041@korea.ac.kr",
17
+ "root": "/workspace/code/CVPR2026/outputs/ablation/re10k/ABLATION_0225_randomSelect",
18
+ "host": "27d18dedec6d",
19
+ "executable": "/venv/main/bin/python",
20
+ "cpu_count": 128,
21
+ "cpu_count_logical": 256,
22
+ "gpu": "NVIDIA H200",
23
+ "gpu_count": 8,
24
+ "disk": {
25
+ "/": {
26
+ "total": "1170378588160",
27
+ "used": "732546121728"
28
+ }
29
+ },
30
+ "memory": {
31
+ "total": "1622948257792"
32
+ },
33
+ "gpu_nvidia": [
34
+ {
35
+ "name": "NVIDIA H200",
36
+ "memoryTotal": "150754820096",
37
+ "cudaCores": 16896,
38
+ "architecture": "Hopper",
39
+ "uuid": "GPU-2649ab80-a3a6-5a1c-0fa5-12bc11bd75e9"
40
+ },
41
+ {
42
+ "name": "NVIDIA H200",
43
+ "memoryTotal": "150754820096",
44
+ "cudaCores": 16896,
45
+ "architecture": "Hopper",
46
+ "uuid": "GPU-e92921d9-c681-246f-af93-637e0dc938ca"
47
+ },
48
+ {
49
+ "name": "NVIDIA H200",
50
+ "memoryTotal": "150754820096",
51
+ "cudaCores": 16896,
52
+ "architecture": "Hopper",
53
+ "uuid": "GPU-ffe12ffc-9bb7-82de-5692-1ec0ee2e68d8"
54
+ },
55
+ {
56
+ "name": "NVIDIA H200",
57
+ "memoryTotal": "150754820096",
58
+ "cudaCores": 16896,
59
+ "architecture": "Hopper",
60
+ "uuid": "GPU-499e5acd-b6ab-2010-c51b-ee9b5aa65825"
61
+ },
62
+ {
63
+ "name": "NVIDIA H200",
64
+ "memoryTotal": "150754820096",
65
+ "cudaCores": 16896,
66
+ "architecture": "Hopper",
67
+ "uuid": "GPU-3b2522d9-1c72-e49b-2c30-96165680b74a"
68
+ },
69
+ {
70
+ "name": "NVIDIA H200",
71
+ "memoryTotal": "150754820096",
72
+ "cudaCores": 16896,
73
+ "architecture": "Hopper",
74
+ "uuid": "GPU-a9a280c5-b2f9-dc1e-a8a9-7326a74001ff"
75
+ },
76
+ {
77
+ "name": "NVIDIA H200",
78
+ "memoryTotal": "150754820096",
79
+ "cudaCores": 16896,
80
+ "architecture": "Hopper",
81
+ "uuid": "GPU-07d0167b-a6a1-1900-2d27-7c6c11598409"
82
+ },
83
+ {
84
+ "name": "NVIDIA H200",
85
+ "memoryTotal": "150754820096",
86
+ "cudaCores": 16896,
87
+ "architecture": "Hopper",
88
+ "uuid": "GPU-8362a999-20d1-c27b-5d18-032d23f859ab"
89
+ }
90
+ ],
91
+ "cudaVersion": "13.1",
92
+ "writerId": "81z9i9czdg5aqgfjdth6b57yriddztva"
93
+ }
ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug-core.log ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {"time":"2026-02-25T10:39:55.558793458Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpxzthjno8/port-145562.txt","pid":145562,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
2
+ {"time":"2026-02-25T10:39:55.559728813Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":145562}
3
+ {"time":"2026-02-25T10:39:55.559688273Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-145562-147994-3073551312/socket","Net":"unix"}}
4
+ {"time":"2026-02-25T10:39:55.732097404Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
5
+ {"time":"2026-02-25T10:39:55.743235764Z","level":"INFO","msg":"handleInformInit: received","streamId":"n76tgbfj","id":"1(@)"}
6
+ {"time":"2026-02-25T10:39:56.464752793Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"n76tgbfj","id":"1(@)"}
7
+ {"time":"2026-02-25T10:40:02.800543333Z","level":"INFO","msg":"connection: cancelling request","id":"1(@)","requestId":"rz66kc5pj9qh"}
ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug-internal.log ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {"time":"2026-02-25T10:39:55.743497828Z","level":"INFO","msg":"stream: starting","core version":"0.25.0"}
2
+ {"time":"2026-02-25T10:39:56.464084002Z","level":"INFO","msg":"stream: created new stream","id":"n76tgbfj"}
3
+ {"time":"2026-02-25T10:39:56.464470748Z","level":"INFO","msg":"handler: started","stream_id":"n76tgbfj"}
4
+ {"time":"2026-02-25T10:39:56.464733133Z","level":"INFO","msg":"stream: started","id":"n76tgbfj"}
5
+ {"time":"2026-02-25T10:39:56.464769993Z","level":"INFO","msg":"sender: started","stream_id":"n76tgbfj"}
6
+ {"time":"2026-02-25T10:39:56.464783063Z","level":"INFO","msg":"writer: started","stream_id":"n76tgbfj"}
ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug.log ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2026-02-25 10:39:55,444 INFO MainThread:145562 [wandb_setup.py:_flush():81] Current SDK version is 0.25.0
2
+ 2026-02-25 10:39:55,444 INFO MainThread:145562 [wandb_setup.py:_flush():81] Configure stats pid to 145562
3
+ 2026-02-25 10:39:55,444 INFO MainThread:145562 [wandb_setup.py:_flush():81] Loading settings from environment variables
4
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:setup_run_log_directory():717] Logging user logs to /workspace/code/CVPR2026/outputs/ablation/re10k/ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug.log
5
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:setup_run_log_directory():718] Logging internal logs to /workspace/code/CVPR2026/outputs/ablation/re10k/ABLATION_0225_randomSelect/wandb/run-20260225_103955-n76tgbfj/logs/debug-internal.log
6
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:init():844] calling init triggers
7
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:init():849] wandb.init called with sweep_config: {}
8
+ config: {'model': {'encoder': {'name': 'dcsplat', 'input_image_shape': [518, 518], 'head_mode': 'pcd', 'num_level': 3, 'gs_param_dim': 256, 'align_corners': False, 'use_voxelize': True}, 'decoder': {'name': 'splatting_cuda', 'background_color': [0.0, 0.0, 0.0], 'make_scale_invariant': False}, 'density_control': {'name': 'density_control_module', 'mean_dim': 32, 'gs_param_dim': 256, 'refinement_layer_num': 1, 'num_level': 3, 'grad_mode': 'absgrad', 'use_mean_features': True, 'refinement_type': 'voxelize', 'refinement_hidden_dim': 32, 'aggregation_mode': 'mean', 'num_heads': 1, 'score_mode': 'random', 'latent_dim': 128, 'num_latents': 64, 'num_self_attn_per_block': 2, 'voxel_size': 0.001, 'aux_refine': False, 'refine_error': False, 'use_refine_module': True, 'voxelize_activate': True, 'use_depth': False}}, 'render_loss': {'mse': {'weight': 1.0}, 'lpips': {'weight': 0.05, 'apply_after_step': 0}}, 'density_control_loss': {'error_score': {'weight': 0.01, 'log_scale': False, 'grad_scale': 10000.0, 'mode': 'original'}}, 'direct_loss': {'l1': {'weight': 0.8}, 'ssim': {'weight': 0.2}}, 'wandb': {'project': 'DCSplat', 'entity': 'scene-representation-group', 'name': 'ABLATION_0225_randomSelect', 'mode': 'online', 'tags': ['re10k', '256x256']}, 'mode': 'train', 'data_loader': {'train': {'num_workers': 16, 'persistent_workers': True, 'batch_size': 16, 'seed': 1234}, 'test': {'num_workers': 4, 'persistent_workers': False, 'batch_size': 1, 'seed': 2345}, 'val': {'num_workers': 1, 'persistent_workers': True, 'batch_size': 1, 'seed': 3456}}, 'optimizer': {'lr': 0.0002, 'warm_up_steps': 25, 'backbone_lr_multiplier': 0.1, 'backbone_trainable': 'T+H', 'accumulate': 1}, 'checkpointing': {'load': None, 'every_n_train_steps': 1500, 'save_top_k': 2, 'save_weights_only': False}, 'train': {'extended_visualization': False, 'print_log_every_n_steps': 10, 'camera_loss': 10.0, 'one_sample_validation': None, 'align_corners': False, 'intrinsic_scaling': False, 'verbose': False, 'beta_dist_param': [0.5, 4.0], 'use_refine_aux': False, 'train_target_set': True, 'train_gs_num': 1, 'ext_scale_detach': False, 'cam_scale_mode': 'sum', 'scene_scale_reg_loss': 0.01, 'train_aux': True, 'vggt_cam_loss': True, 'vggt_distil': False, 'context_view_train': False}, 'test': {'output_path': 'test/ablation/re10k', 'align_pose': False, 'pose_align_steps': 100, 'rot_opt_lr': 0.005, 'trans_opt_lr': 0.005, 'compute_scores': True, 'save_image': False, 'save_video': False, 'save_active_mask_image': False, 'save_error_score_image': False, 'save_compare': False, 'pred_intrinsic': False, 'error_threshold': 0.4, 'error_threshold_list': [0.2, 0.4, 0.6, 0.8, 1.0], 'threshold_mode': 'ratio', 'nvs_view_N_list': [3, 6, 16, 32, 64]}, 'seed': 111123, 'trainer': {'max_steps': 3001, 'val_check_interval': 250, 'gradient_clip_val': 0.5, 'num_nodes': 1}, 'dataset': {'re10k': {'make_baseline_1': True, 'relative_pose': True, 'augment': True, 'background_color': [0.0, 0.0, 0.0], 'overfit_to_scene': None, 'skip_bad_shape': True, 'view_sampler': {'name': 'bounded', 'num_target_views': 4, 'num_context_views': 2, 'min_distance_between_context_views': 45, 'max_distance_between_context_views': 90, 'min_distance_to_context_views': 0, 'warm_up_steps': 1000, 'initial_min_distance_between_context_views': 25, 'initial_max_distance_between_context_views': 25, 'same_target_gap': False, 'num_target_set': 3}, 'name': 're10k', 'roots': ['datasets/re10k'], 'input_image_shape': [256, 256], 'original_image_shape': [360, 640], 'cameras_are_circular': False, 'baseline_min': 0.001, 'baseline_max': 10000000000.0, 'max_fov': 100.0, 'dynamic_context_views': True, 'max_context_views_per_gpu': 24}}, '_wandb': {}}
9
+ 2026-02-25 10:39:55,445 INFO MainThread:145562 [wandb_init.py:init():892] starting backend
10
+ 2026-02-25 10:39:55,732 INFO MainThread:145562 [wandb_init.py:init():895] sending inform_init request
11
+ 2026-02-25 10:39:55,739 INFO MainThread:145562 [wandb_init.py:init():903] backend started and connected
12
+ 2026-02-25 10:39:55,744 INFO MainThread:145562 [wandb_init.py:init():973] updated telemetry
13
+ 2026-02-25 10:39:55,752 INFO MainThread:145562 [wandb_init.py:init():997] communicating run to backend with 90.0 second timeout
14
+ 2026-02-25 10:39:57,667 INFO MainThread:145562 [wandb_init.py:init():1042] starting run threads in backend
15
+ 2026-02-25 10:39:57,796 INFO MainThread:145562 [wandb_run.py:_console_start():2524] atexit reg
16
+ 2026-02-25 10:39:57,797 INFO MainThread:145562 [wandb_run.py:_redirect():2373] redirect: wrap_raw
17
+ 2026-02-25 10:39:57,797 INFO MainThread:145562 [wandb_run.py:_redirect():2442] Wrapping output streams.
18
+ 2026-02-25 10:39:57,797 INFO MainThread:145562 [wandb_run.py:_redirect():2465] Redirects installed.
19
+ 2026-02-25 10:39:57,800 INFO MainThread:145562 [wandb_init.py:init():1082] run started, returning control to user process
ABLATION_0225_targetTrain/main.log ADDED
@@ -0,0 +1,128 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 01:40:07,061][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 01:40:14,380][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 01:40:14,380][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 01:41:04,283][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:425: The 'val_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=31` in the `DataLoader` to improve performance.
9
+
10
+ [2026-02-25 01:41:04,285][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
11
+ warnings.warn( # warn only once
12
+
13
+ [2026-02-25 01:41:06,636][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
14
+ result[selector] = overlay
15
+
16
+ [2026-02-25 01:41:06,645][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/utilities/data.py:79: Trying to infer the `batch_size` from an ambiguous collection. The batch size we found is 1. To avoid any miscalculations, use `self.log(..., batch_size=batch_size)`.
17
+
18
+ [2026-02-25 01:41:06,646][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
19
+ warnings.warn(
20
+
21
+ [2026-02-25 01:41:06,646][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
22
+ warnings.warn(msg)
23
+
24
+ [2026-02-25 01:41:08,321][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/functional.py:554: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /pytorch/aten/src/ATen/native/TensorShape.cpp:4322.)
25
+ return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]
26
+
27
+ [2026-02-25 01:41:08,603][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/psnr', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
28
+
29
+ [2026-02-25 01:41:08,605][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/lpips', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
30
+
31
+ [2026-02-25 01:41:08,605][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/ssim', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
32
+
33
+ [2026-02-25 01:41:08,605][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/gaussian_num_ratio', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
34
+
35
+ [2026-02-25 01:41:08,605][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('info/global_step', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
36
+
37
+ [2026-02-25 01:41:16,604][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
38
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
39
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
40
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
41
+
42
+ [2026-02-25 01:41:16,706][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
43
+ result[selector] = overlay
44
+
45
+ [2026-02-25 01:42:41,674][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
46
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
47
+
48
+ [2026-02-25 01:52:46,552][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
49
+ result[selector] = overlay
50
+
51
+ [2026-02-25 01:55:34,591][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
52
+ result[selector] = overlay
53
+
54
+ [2026-02-25 02:04:17,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
55
+ result[selector] = overlay
56
+
57
+ [2026-02-25 02:09:55,959][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
58
+ result[selector] = overlay
59
+
60
+ [2026-02-25 02:15:44,850][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
61
+ result[selector] = overlay
62
+
63
+ [2026-02-25 02:24:20,918][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
64
+ result[selector] = overlay
65
+
66
+ [2026-02-25 02:27:18,595][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
67
+ result[selector] = overlay
68
+
69
+ [2026-02-25 02:38:48,014][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
70
+ result[selector] = overlay
71
+
72
+ [2026-02-25 02:38:52,181][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
73
+ result[selector] = overlay
74
+
75
+ [2026-02-25 02:50:26,889][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
76
+ result[selector] = overlay
77
+
78
+ [2026-02-25 02:53:18,593][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
79
+ result[selector] = overlay
80
+
81
+ [2026-02-25 03:02:02,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
82
+ result[selector] = overlay
83
+
84
+ [2026-02-25 03:07:58,736][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
85
+ result[selector] = overlay
86
+
87
+ [2026-02-25 03:13:49,838][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
88
+ result[selector] = overlay
89
+
90
+ [2026-02-25 03:22:26,000][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
91
+ result[selector] = overlay
92
+
93
+ [2026-02-25 03:25:22,744][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
94
+ result[selector] = overlay
95
+
96
+ [2026-02-25 03:36:53,441][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
97
+ result[selector] = overlay
98
+
99
+ [2026-02-25 03:36:56,656][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
100
+ result[selector] = overlay
101
+
102
+ [2026-02-25 03:48:29,358][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
103
+ result[selector] = overlay
104
+
105
+ [2026-02-25 03:51:23,880][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
106
+ result[selector] = overlay
107
+
108
+ [2026-02-25 04:00:10,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
109
+ result[selector] = overlay
110
+
111
+ [2026-02-25 04:05:52,901][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
112
+ result[selector] = overlay
113
+
114
+ [2026-02-25 04:11:47,470][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
115
+ result[selector] = overlay
116
+
117
+ [2026-02-25 04:20:24,888][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
118
+ result[selector] = overlay
119
+
120
+ [2026-02-25 04:23:24,966][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
121
+ result[selector] = overlay
122
+
123
+ [2026-02-25 04:35:15,500][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
124
+ result[selector] = overlay
125
+
126
+ [2026-02-25 04:35:19,661][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
127
+ result[selector] = overlay
128
+
ABLATION_0225_targetTrain/peak_vram_memory.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "peak_memory_allocated_gb": 78.034,
3
+ "peak_memory_reserved_gb": 137.805,
4
+ "total_elapsed_hours": 2.92,
5
+ "mode": "train"
6
+ }
ABLATION_0225_targetTrain/train_ddp_process_1.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 01:40:24,769][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 01:40:51,857][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 01:40:51,857][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 01:41:04,284][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 01:41:16,606][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 01:41:16,725][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 01:42:41,650][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 01:52:46,552][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 02:04:17,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 02:15:44,851][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 02:27:18,595][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 02:38:52,179][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 02:50:26,889][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 03:02:02,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 03:13:49,838][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 03:25:22,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 03:36:56,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 03:48:29,357][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 04:00:10,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 04:11:47,469][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 04:23:24,966][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 04:35:19,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain/train_ddp_process_2.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 01:40:24,794][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 01:40:51,561][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 01:40:51,562][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 01:41:04,284][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 01:41:16,606][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 01:41:16,718][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 01:42:41,673][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 01:52:46,553][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 02:04:17,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 02:15:44,850][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 02:27:18,595][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 02:38:52,179][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 02:50:26,889][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 03:02:02,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 03:13:49,837][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 03:25:22,743][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 03:36:56,656][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 03:48:29,357][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 04:00:10,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 04:11:47,470][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 04:23:24,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 04:35:19,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain/train_ddp_process_3.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 01:40:24,827][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 01:40:51,728][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 01:40:51,728][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 01:41:04,284][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 01:41:16,607][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 01:41:16,716][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 01:42:41,649][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 01:52:46,552][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 02:04:17,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 02:15:44,849][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 02:27:18,594][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 02:38:52,179][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 02:50:26,889][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 03:02:02,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 03:13:49,837][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 03:25:22,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 03:36:56,656][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 03:48:29,357][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 04:00:10,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 04:11:47,469][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 04:23:24,966][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 04:35:19,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain/train_ddp_process_4.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 01:40:24,824][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 01:40:51,803][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 01:40:51,803][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 01:41:04,284][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 01:41:16,602][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 01:41:16,711][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 01:42:41,649][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 01:52:46,552][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 02:04:17,314][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 02:15:44,850][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 02:27:18,595][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 02:38:52,179][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 02:50:26,891][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 03:02:02,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 03:13:49,838][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 03:25:22,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 03:36:56,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 03:48:29,357][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 04:00:10,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 04:11:47,470][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 04:23:24,966][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 04:35:19,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain/train_ddp_process_5.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 01:40:24,808][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 01:40:40,707][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 01:40:40,708][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 01:41:04,284][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 01:41:15,863][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 01:41:16,724][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 01:42:41,649][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 01:52:46,552][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 02:04:17,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 02:15:44,849][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 02:27:18,594][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 02:38:52,179][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 02:50:26,889][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 03:02:02,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 03:13:49,837][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 03:25:22,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 03:36:56,656][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 03:48:29,356][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 04:00:10,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 04:11:47,469][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 04:23:24,966][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 04:35:19,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain/train_ddp_process_6.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 01:40:24,809][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 01:40:41,182][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 01:40:41,182][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 01:41:04,284][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 01:41:16,088][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 01:41:16,713][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 01:42:41,677][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 01:52:46,552][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 02:04:17,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 02:15:44,850][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 02:27:18,594][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 02:38:52,179][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 02:50:26,889][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 03:02:02,968][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 03:13:49,837][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 03:25:22,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 03:36:56,656][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 03:48:29,356][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 04:00:10,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 04:11:47,469][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 04:23:24,966][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 04:35:19,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain/train_ddp_process_7.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 01:40:24,726][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 01:40:51,954][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 01:40:51,954][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 01:41:04,284][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 01:41:16,101][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 01:41:16,714][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 01:42:41,669][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 01:52:46,552][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 02:04:17,312][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 02:15:44,850][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 02:27:18,594][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 02:38:52,179][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 02:50:26,889][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 03:02:02,967][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 03:13:49,839][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 03:25:22,742][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 03:36:56,656][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 03:48:29,357][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 04:00:10,015][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 04:11:47,469][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 04:23:24,966][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 04:35:19,657][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain_SSR/main.log ADDED
@@ -0,0 +1,128 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 04:35:40,752][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 04:35:47,030][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 04:35:47,030][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 04:36:41,641][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:425: The 'val_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=31` in the `DataLoader` to improve performance.
9
+
10
+ [2026-02-25 04:36:41,642][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
11
+ warnings.warn( # warn only once
12
+
13
+ [2026-02-25 04:36:44,110][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
14
+ result[selector] = overlay
15
+
16
+ [2026-02-25 04:36:44,120][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/utilities/data.py:79: Trying to infer the `batch_size` from an ambiguous collection. The batch size we found is 1. To avoid any miscalculations, use `self.log(..., batch_size=batch_size)`.
17
+
18
+ [2026-02-25 04:36:44,120][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
19
+ warnings.warn(
20
+
21
+ [2026-02-25 04:36:44,121][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
22
+ warnings.warn(msg)
23
+
24
+ [2026-02-25 04:36:45,824][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/functional.py:554: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /pytorch/aten/src/ATen/native/TensorShape.cpp:4322.)
25
+ return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]
26
+
27
+ [2026-02-25 04:36:46,108][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/psnr', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
28
+
29
+ [2026-02-25 04:36:46,109][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/lpips', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
30
+
31
+ [2026-02-25 04:36:46,110][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/ssim', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
32
+
33
+ [2026-02-25 04:36:46,110][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('val/gaussian_num_ratio', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
34
+
35
+ [2026-02-25 04:36:46,110][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/lightning/pytorch/trainer/connectors/logger_connector/result.py:434: It is recommended to use `self.log('info/global_step', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
36
+
37
+ [2026-02-25 04:36:54,723][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
38
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
39
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
40
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
41
+
42
+ [2026-02-25 04:36:54,811][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
43
+ result[selector] = overlay
44
+
45
+ [2026-02-25 04:38:19,468][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
46
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
47
+
48
+ [2026-02-25 04:48:25,988][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
49
+ result[selector] = overlay
50
+
51
+ [2026-02-25 04:51:14,127][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
52
+ result[selector] = overlay
53
+
54
+ [2026-02-25 05:00:01,620][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
55
+ result[selector] = overlay
56
+
57
+ [2026-02-25 05:05:41,307][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
58
+ result[selector] = overlay
59
+
60
+ [2026-02-25 05:11:31,315][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
61
+ result[selector] = overlay
62
+
63
+ [2026-02-25 05:20:07,059][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
64
+ result[selector] = overlay
65
+
66
+ [2026-02-25 05:23:05,955][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
67
+ result[selector] = overlay
68
+
69
+ [2026-02-25 05:34:35,366][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
70
+ result[selector] = overlay
71
+
72
+ [2026-02-25 05:34:39,776][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
73
+ result[selector] = overlay
74
+
75
+ [2026-02-25 05:46:15,972][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
76
+ result[selector] = overlay
77
+
78
+ [2026-02-25 05:49:06,959][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
79
+ result[selector] = overlay
80
+
81
+ [2026-02-25 05:57:53,153][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
82
+ result[selector] = overlay
83
+
84
+ [2026-02-25 06:03:52,411][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
85
+ result[selector] = overlay
86
+
87
+ [2026-02-25 06:09:43,025][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
88
+ result[selector] = overlay
89
+
90
+ [2026-02-25 06:18:21,029][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
91
+ result[selector] = overlay
92
+
93
+ [2026-02-25 06:21:18,231][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
94
+ result[selector] = overlay
95
+
96
+ [2026-02-25 06:32:50,731][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
97
+ result[selector] = overlay
98
+
99
+ [2026-02-25 06:32:53,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
100
+ result[selector] = overlay
101
+
102
+ [2026-02-25 06:44:28,349][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
103
+ result[selector] = overlay
104
+
105
+ [2026-02-25 06:47:21,393][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
106
+ result[selector] = overlay
107
+
108
+ [2026-02-25 06:56:07,075][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
109
+ result[selector] = overlay
110
+
111
+ [2026-02-25 07:01:50,987][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
112
+ result[selector] = overlay
113
+
114
+ [2026-02-25 07:07:47,162][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
115
+ result[selector] = overlay
116
+
117
+ [2026-02-25 07:16:25,080][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
118
+ result[selector] = overlay
119
+
120
+ [2026-02-25 07:19:25,492][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
121
+ result[selector] = overlay
122
+
123
+ [2026-02-25 07:31:11,035][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
124
+ result[selector] = overlay
125
+
126
+ [2026-02-25 07:31:15,205][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
127
+ result[selector] = overlay
128
+
ABLATION_0225_targetTrain_SSR/peak_vram_memory.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "peak_memory_allocated_gb": 78.07,
3
+ "peak_memory_reserved_gb": 132.943,
4
+ "total_elapsed_hours": 2.93,
5
+ "mode": "train"
6
+ }
ABLATION_0225_targetTrain_SSR/train_ddp_process_1.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 04:35:57,761][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 04:36:27,539][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 04:36:27,539][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 04:36:41,642][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 04:36:54,719][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 04:36:54,847][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 04:38:19,450][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 04:48:25,988][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 05:00:01,620][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 05:11:31,315][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 05:23:05,955][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 05:34:39,775][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 05:46:15,971][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 05:57:53,153][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 06:09:43,025][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 06:21:18,231][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 06:32:53,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 06:44:28,348][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 06:56:07,074][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 07:07:47,162][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 07:19:25,492][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 07:31:15,203][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain_SSR/train_ddp_process_3.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 04:35:57,702][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 04:36:24,380][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 04:36:24,384][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 04:36:41,642][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 04:36:54,719][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 04:36:54,946][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 04:38:19,450][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 04:48:25,988][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 05:00:01,620][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 05:11:31,315][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 05:23:05,955][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 05:34:39,775][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 05:46:15,971][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 05:57:53,153][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 06:09:43,025][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 06:21:18,231][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 06:32:53,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 06:44:28,348][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 06:56:07,074][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 07:07:47,163][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 07:19:25,492][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 07:31:15,203][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain_SSR/train_ddp_process_5.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 04:35:57,766][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 04:36:11,387][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 04:36:11,387][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 04:36:41,642][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 04:36:53,976][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 04:36:54,822][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 04:38:19,450][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 04:48:25,988][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 05:00:01,621][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 05:11:31,315][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 05:23:05,955][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 05:34:39,775][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 05:46:15,971][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 05:57:53,297][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 06:09:43,025][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 06:21:18,231][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 06:32:53,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 06:44:28,348][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 06:56:07,074][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 07:07:47,162][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 07:19:25,492][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 07:31:15,203][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+
ABLATION_0225_targetTrain_SSR/train_ddp_process_6.log ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2026-02-25 04:35:57,786][dinov2][INFO] - using MLP layer as FFN
2
+ [2026-02-25 04:36:26,541][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.
3
+ warnings.warn(
4
+
5
+ [2026-02-25 04:36:26,541][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
6
+ warnings.warn(msg)
7
+
8
+ [2026-02-25 04:36:41,642][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4807: UserWarning: No device id is provided via `init_process_group` or `barrier `. Using the current device set by the user.
9
+ warnings.warn( # warn only once
10
+
11
+ [2026-02-25 04:36:54,211][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/autograd/graph.py:829: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
12
+ grad.sizes() = [57, 256, 1, 1], strides() = [256, 1, 256, 256]
13
+ bucket_view.sizes() = [57, 256, 1, 1], strides() = [256, 1, 1, 1] (Triggered internally at /pytorch/torch/csrc/distributed/c10d/reducer.cpp:334.)
14
+ return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
15
+
16
+ [2026-02-25 04:36:54,902][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
17
+ result[selector] = overlay
18
+
19
+ [2026-02-25 04:38:19,467][py.warnings][WARNING] - /venv/main/lib/python3.12/site-packages/torch/optim/lr_scheduler.py:209: UserWarning: The epoch parameter in `scheduler.step()` was not necessary and is being deprecated where possible. Please use `scheduler.step()` to step the scheduler. During the deprecation, if epoch is different from None, the closed form is used instead of the new chainable form, where available. Please open an issue if you are unable to replicate your use case: https://github.com/pytorch/pytorch/issues/new/choose.
20
+ warnings.warn(EPOCH_DEPRECATION_WARNING, UserWarning)
21
+
22
+ [2026-02-25 04:48:25,988][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
23
+ result[selector] = overlay
24
+
25
+ [2026-02-25 05:00:01,621][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
26
+ result[selector] = overlay
27
+
28
+ [2026-02-25 05:11:31,315][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
29
+ result[selector] = overlay
30
+
31
+ [2026-02-25 05:23:05,954][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
32
+ result[selector] = overlay
33
+
34
+ [2026-02-25 05:34:39,775][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
35
+ result[selector] = overlay
36
+
37
+ [2026-02-25 05:46:15,971][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
38
+ result[selector] = overlay
39
+
40
+ [2026-02-25 05:57:53,156][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
41
+ result[selector] = overlay
42
+
43
+ [2026-02-25 06:09:43,024][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
44
+ result[selector] = overlay
45
+
46
+ [2026-02-25 06:21:18,230][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
47
+ result[selector] = overlay
48
+
49
+ [2026-02-25 06:32:53,993][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
50
+ result[selector] = overlay
51
+
52
+ [2026-02-25 06:44:28,348][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
53
+ result[selector] = overlay
54
+
55
+ [2026-02-25 06:56:07,074][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
56
+ result[selector] = overlay
57
+
58
+ [2026-02-25 07:07:47,162][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
59
+ result[selector] = overlay
60
+
61
+ [2026-02-25 07:19:25,492][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
62
+ result[selector] = overlay
63
+
64
+ [2026-02-25 07:31:15,203][py.warnings][WARNING] - /workspace/code/CVPR2026/src/visualization/layout.py:105: UserWarning: Using a non-tuple sequence for multidimensional indexing is deprecated and will be changed in pytorch 2.9; use x[tuple(seq)] instead of x[seq]. In pytorch 2.9 this will be interpreted as tensor index, x[torch.tensor(seq)], which will result either in an error or a different result (Triggered internally at /pytorch/torch/csrc/autograd/python_variable_indexing.cpp:316.)
65
+ result[selector] = overlay
66
+