SnoopyFan commited on
Commit
8fb5b54
·
verified ·
1 Parent(s): c7f6fd5

Add WorldCompass Sana training artifacts

Browse files
rebuttal_data/worldcompass_sana/wc_sana_smoke_82b1191_20260801d/node_0/torchrun.log ADDED
@@ -0,0 +1,158 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ /mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/_distutils_hack/__init__.py:53: UserWarning: Reliance on distutils from stdlib is deprecated. Users must rely on setuptools to provide the distutils module. Avoid importing distutils or import setuptools first, and avoid setting SETUPTOOLS_USE_DISTUTILS=stdlib. Register concerns at https://github.com/pypa/setuptools/issues/new?template=distutils-deprecation.yml
2
+ warnings.warn(
3
+ /mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/_distutils_hack/__init__.py:53: UserWarning: Reliance on distutils from stdlib is deprecated. Users must rely on setuptools to provide the distutils module. Avoid importing distutils or import setuptools first, and avoid setting SETUPTOOLS_USE_DISTUTILS=stdlib. Register concerns at https://github.com/pypa/setuptools/issues/new?template=distutils-deprecation.yml
4
+ warnings.warn(
5
+ /mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/mmcv/__init__.py:20: UserWarning: On January 1, 2023, MMCV will release v2.0.0, in which it will remove components related to the training process and add a data transformation module. In addition, it will rename the package names mmcv to mmcv-lite and mmcv-full to mmcv. See https://github.com/open-mmlab/mmcv/blob/master/docs/en/compatibility.md for more details.
6
+ warnings.warn(
7
+ /mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/mmcv/__init__.py:20: UserWarning: On January 1, 2023, MMCV will release v2.0.0, in which it will remove components related to the training process and add a data transformation module. In addition, it will rename the package names mmcv to mmcv-lite and mmcv-full to mmcv. See https://github.com/open-mmlab/mmcv/blob/master/docs/en/compatibility.md for more details.
8
+ warnings.warn(
9
+ /mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/wan/model.py:124: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
10
+ @amp.autocast(enabled=False)
11
+ /mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/wan/model.py:124: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
12
+ @amp.autocast(enabled=False)
13
+ /mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/wan/model.py:134: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
14
+ @amp.autocast(enabled=False)
15
+ /mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/wan/model.py:134: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
16
+ @amp.autocast(enabled=False)
17
+ /mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/nets/fastlinear/modules/nn/norm.py:186: UserWarning: Cannot import apex RMSNorm, switch to vanilla implementation
18
+ warnings.warn("Cannot import apex RMSNorm, switch to vanilla implementation")
19
+ /mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/nets/fastlinear/modules/nn/norm.py:186: UserWarning: Cannot import apex RMSNorm, switch to vanilla implementation
20
+ warnings.warn("Cannot import apex RMSNorm, switch to vanilla implementation")
21
+ 2026-08-02 00:57:53 - [Sana] - INFO - [FSDP2] block=SanaVideoMSCamCtrlBlock, cp_size=1, reshard_after_forward=True, limit_all_gathers=False, backward_prefetch=backward_pre
22
+ backward_prefetch is not supported in FSDP2. Setting backward prefetch to None.
23
+ backward_prefetch is not supported in FSDP2. Setting backward prefetch to None.
24
+ 2026-08-02 00:57:57 - [Sana] - INFO - Failed to load null embed from None....'NoneType' object has no attribute 'seek'. You can only torch.load from a file that is seekable. Please pre-load the data into a buffer like io.BytesIO and try to load from it instead.. Ignore the error during inference
25
+ 2026-08-02 00:57:57 - [Sana] - INFO - use pe: True, pos embed type: casual_wan_rope, position embed interpolation: 1.0, base size: 22
26
+ 2026-08-02 00:57:57 - [Sana] - INFO - attention type: BidirectionalGDNTriton; ffn type: CachedGLUMBConvTemp; self-attn head dim: 112; self-attn qk norm: True; cross-attn type: flash; cross-attn qk norm: True; autocast linear attn: False
27
+ 2026-08-02 00:57:58 - [Sana] - INFO - Hybrid attention (softmax_every_n=4):
28
+ attn_type_list = ['BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton', 'BidirectionalGDNTriton']
29
+ camctrl_cls_list = ['CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedSoftmaxUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedSoftmaxUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedSoftmaxUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedSoftmaxUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedChunkCausalGDNUCPESinglePathLiteLA', 'CachedSoftmaxUCPESinglePathLiteLA']
30
+ 2026-08-02 00:58:15 - [Sana] - INFO - Failed to load null embed from None....'NoneType' object has no attribute 'seek'. You can only torch.load from a file that is seekable. Please pre-load the data into a buffer like io.BytesIO and try to load from it instead.. Ignore the error during inference
31
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
32
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
33
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
34
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
35
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
36
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
37
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
38
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
39
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
40
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
41
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
42
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
43
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
44
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
45
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
46
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
47
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
48
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
49
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
50
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
51
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
52
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
53
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
54
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA[Sana] Loading model from /mnt/workspace/fly/reward_dmd_sanawm/models/SANA-WM_streaming/fb4ef609f5da47d5815e850a5ecec9a12f069b57/sana_dit/model.pt
55
+
56
+ [Sana] Loaded model from /mnt/workspace/fly/reward_dmd_sanawm/models/SANA-WM_streaming/fb4ef609f5da47d5815e850a5ecec9a12f069b57/sana_dit/model.pt
57
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
58
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
59
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
60
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
61
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
62
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
63
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
64
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
65
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
66
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
67
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
68
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
69
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
70
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
71
+ Initializing camera branch QKV from base model QKV for CachedChunkCausalGDNUCPESinglePathLiteLA
72
+ Initializing camera branch QKV from base model QKV for CachedSoftmaxUCPESinglePathLiteLA
73
+ [Sana] Loading model from /mnt/workspace/fly/reward_dmd_sanawm/models/SANA-WM_streaming/fb4ef609f5da47d5815e850a5ecec9a12f069b57/sana_dit/model.pt
74
+ [Sana] Loaded model from /mnt/workspace/fly/reward_dmd_sanawm/models/SANA-WM_streaming/fb4ef609f5da47d5815e850a5ecec9a12f069b57/sana_dit/model.pt
75
+ 2026-08-02 00:59:06 - [Sana] - INFO - Loaded /mnt/workspace/fly/reward_dmd_sanawm/models/SANA-WM_streaming/fb4ef609f5da47d5815e850a5ecec9a12f069b57/sana_dit/model.pt: model=SanaMSVideoCamCtrlStreaming_1600M_P1_D20, missing=1, unexpected=0
76
+ 2026-08-02 00:59:09 - [Sana] - INFO - FSDP2: removing custom __call__ from SanaMSVideoCamCtrlStreaming
77
+ 2026-08-02 00:59:09 - [Sana] - INFO - FSDP2: removing custom __call__ from Sana
78
+ [rank0]: Traceback (most recent call last):
79
+ [rank0]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/train_video_scripts/train_longsana.py", line 591, in <module>
80
+ [rank0]: main()
81
+ [rank0]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/train_video_scripts/train_longsana.py", line 570, in main
82
+ [rank0]: trainer = SanaWMWorldCompassTrainer(config)
83
+ [rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
84
+ [rank0]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/longsana/trainer/sana_wm_worldcompass.py", line 349, in __init__
85
+ [rank0]: SanaWMDistillTrainer.__init__(
86
+ [rank0]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/longsana/trainer/sana_wm_distill.py", line 358, in __init__
87
+ [rank0]: self.text_encoder = SanaTextEncoder(
88
+ [rank0]: ^^^^^^^^^^^^^^^^
89
+ [rank0]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/longsana/utils/model_wrapper.py", line 174, in __init__
90
+ [rank0]: self.tokenizer, self.text_encoder = get_tokenizer_and_text_encoder(name=name, device=device)
91
+ [rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
92
+ [rank0]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/builder.py", line 113, in get_tokenizer_and_text_encoder
93
+ [rank0]: pretrained_source, local_files_only = _resolve_text_encoder_source(
94
+ [rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
95
+ [rank0]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/builder.py", line 70, in _resolve_text_encoder_source
96
+ [rank0]: raise RuntimeError(
97
+ [rank0]: RuntimeError: SANA_GEMMA_2_2B_IT_PATH must pin an absolute local snapshot while offline mode is enabled
98
+ 2026-08-02 00:59:09 - [Sana] - INFO - FSDP2: removing custom __call__ from SanaMSVideoCamCtrlStreaming
99
+ 2026-08-02 00:59:09 - [Sana] - INFO - FSDP2: removing custom __call__ from Sana
100
+ [rank1]: Traceback (most recent call last):
101
+ [rank1]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/train_video_scripts/train_longsana.py", line 591, in <module>
102
+ [rank1]: main()
103
+ [rank1]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/train_video_scripts/train_longsana.py", line 570, in main
104
+ [rank1]: trainer = SanaWMWorldCompassTrainer(config)
105
+ [rank1]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
106
+ [rank1]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/longsana/trainer/sana_wm_worldcompass.py", line 349, in __init__
107
+ [rank1]: SanaWMDistillTrainer.__init__(
108
+ [rank1]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/longsana/trainer/sana_wm_distill.py", line 358, in __init__
109
+ [rank1]: self.text_encoder = SanaTextEncoder(
110
+ [rank1]: ^^^^^^^^^^^^^^^^
111
+ [rank1]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/longsana/utils/model_wrapper.py", line 174, in __init__
112
+ [rank1]: self.tokenizer, self.text_encoder = get_tokenizer_and_text_encoder(name=name, device=device)
113
+ [rank1]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
114
+ [rank1]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/builder.py", line 113, in get_tokenizer_and_text_encoder
115
+ [rank1]: pretrained_source, local_files_only = _resolve_text_encoder_source(
116
+ [rank1]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
117
+ [rank1]: File "/mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/diffusion/model/builder.py", line 70, in _resolve_text_encoder_source
118
+ [rank1]: raise RuntimeError(
119
+ [rank1]: RuntimeError: SANA_GEMMA_2_2B_IT_PATH must pin an absolute local snapshot while offline mode is enabled
120
+ [rank0]:[W802 00:59:10.469779550 ProcessGroupNCCL.cpp:1524] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator())
121
+ [rank0]:[W802 00:59:11.314840219 AllocatorConfig.cpp:28] Warning: PYTORCH_CUDA_ALLOC_CONF is deprecated, use PYTORCH_ALLOC_CONF instead (function operator())
122
+ [rank1]:[W802 00:59:11.683124990 AllocatorConfig.cpp:28] Warning: PYTORCH_CUDA_ALLOC_CONF is deprecated, use PYTORCH_ALLOC_CONF instead (function operator())
123
+ W0802 00:59:12.277000 295818 torch/distributed/elastic/multiprocessing/api.py:908] Sending process 295833 closing signal SIGTERM
124
+ E0802 00:59:12.343000 295818 torch/distributed/elastic/multiprocessing/api.py:882] failed (exitcode: 1) local_rank: 0 (pid: 295832) of binary: /mnt/workspace/fly/reward_dmd_sanawm/.venv/bin/python
125
+ Traceback (most recent call last):
126
+ File "<frozen runpy>", line 198, in _run_module_as_main
127
+ File "<frozen runpy>", line 88, in _run_code
128
+ File "/mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/torch/distributed/run.py", line 940, in <module>
129
+ main()
130
+ File "/mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 357, in wrapper
131
+ return f(*args, **kwargs)
132
+ ^^^^^^^^^^^^^^^^^^
133
+ File "/mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/torch/distributed/run.py", line 936, in main
134
+ run(args)
135
+ File "/mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/torch/distributed/run.py", line 927, in run
136
+ elastic_launch(
137
+ File "/mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/torch/distributed/launcher/api.py", line 156, in __call__
138
+ return launch_agent(self._config, self._entrypoint, list(args))
139
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
140
+ File "/mnt/workspace/fly/reward_dmd_sanawm/.venv/lib/python3.11/site-packages/torch/distributed/launcher/api.py", line 293, in launch_agent
141
+ raise ChildFailedError(
142
+ torch.distributed.elastic.multiprocessing.errors.ChildFailedError:
143
+ ============================================================
144
+ /mnt/workspace/fly/reward_dmd_sanawm/Sana_RF_24card/train_video_scripts/train_longsana.py FAILED
145
+ ------------------------------------------------------------
146
+ Failures:
147
+ <NO_OTHER_FAILURES>
148
+ ------------------------------------------------------------
149
+ Root Cause (first observed failure):
150
+ [0]:
151
+ time : 2026-08-02_00:59:12
152
+ host : dsw-80423-5665c9b68b-klz7l
153
+ rank : 0 (local_rank: 0)
154
+ exitcode : 1 (pid: 295832)
155
+ error_file: <N/A>
156
+ traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html
157
+ ============================================================
158
+ [W802 00:59:12.352115414 AllocatorConfig.cpp:28] Warning: PYTORCH_CUDA_ALLOC_CONF is deprecated, use PYTORCH_ALLOC_CONF instead (function operator())