Hugging Face's logo

htlou
/
wandb

Model card Files
xet
Community
htlou commited on
Commit
befabe2
·
verified ·
1 Parent(s): f5d339e

Upload folder using huggingface_hub

Browse files
Files changed (17) hide show
  1. .gitattributes +2 -0
  2. debug-internal.log +9 -7
  3. debug.log +22 -26
  4. run-20250401_170224-799gcj60/files/output.log +0 -0
  5. run-20250401_170224-799gcj60/files/requirements.txt +258 -0
  6. run-20250401_170224-799gcj60/files/wandb-metadata.json +106 -0
  7. run-20250401_170224-799gcj60/logs/debug-core.log +7 -0
  8. run-20250401_170224-799gcj60/logs/debug-internal.log +7 -0
  9. run-20250401_170224-799gcj60/logs/debug.log +22 -0
  10. run-20250401_170224-799gcj60/run-799gcj60.wandb +3 -0
  11. run-20250401_204801-q89zozii/files/output.log +0 -0
  12. run-20250401_204801-q89zozii/files/requirements.txt +258 -0
  13. run-20250401_204801-q89zozii/files/wandb-metadata.json +106 -0
  14. run-20250401_204801-q89zozii/logs/debug-core.log +7 -0
  15. run-20250401_204801-q89zozii/logs/debug-internal.log +9 -0
  16. run-20250401_204801-q89zozii/logs/debug.log +22 -0
  17. run-20250401_204801-q89zozii/run-q89zozii.wandb +3 -0
.gitattributes CHANGED
@@ -37,3 +37,5 @@ run-20250329_010934-3x35hjks/run-3x35hjks.wandb filter=lfs diff=lfs merge=lfs -t
37
  run-20250329_012205-co1ecmky/run-co1ecmky.wandb filter=lfs diff=lfs merge=lfs -text
38
  offline-run-20250328_203038-ni3gydc1/run-ni3gydc1.wandb filter=lfs diff=lfs merge=lfs -text
39
  offline-run-20250328_203844-gc2qytpj/run-gc2qytpj.wandb filter=lfs diff=lfs merge=lfs -text
 
 
 
37
  run-20250329_012205-co1ecmky/run-co1ecmky.wandb filter=lfs diff=lfs merge=lfs -text
38
  offline-run-20250328_203038-ni3gydc1/run-ni3gydc1.wandb filter=lfs diff=lfs merge=lfs -text
39
  offline-run-20250328_203844-gc2qytpj/run-gc2qytpj.wandb filter=lfs diff=lfs merge=lfs -text
40
+ run-20250401_170224-799gcj60/run-799gcj60.wandb filter=lfs diff=lfs merge=lfs -text
41
+ run-20250401_204801-q89zozii/run-q89zozii.wandb filter=lfs diff=lfs merge=lfs -text
debug-internal.log CHANGED
@@ -1,7 +1,9 @@
1
- {"time":"2025-03-28T20:38:44.692258573+08:00","level":"INFO","msg":"stream: starting","core version":"0.19.8","symlink path":"../outputs/test_13B/wandb/offline-run-20250328_203844-gc2qytpj/logs/debug-core.log"}
2
- {"time":"2025-03-28T20:38:44.90653769+08:00","level":"INFO","msg":"created new stream","id":"gc2qytpj"}
3
- {"time":"2025-03-28T20:38:44.90660149+08:00","level":"INFO","msg":"stream: started","id":"gc2qytpj"}
4
- {"time":"2025-03-28T20:38:44.906615038+08:00","level":"INFO","msg":"handler: started","stream_id":"gc2qytpj"}
5
- {"time":"2025-03-28T20:38:44.906633311+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"gc2qytpj"}
6
- {"time":"2025-03-28T20:38:44.90663431+08:00","level":"INFO","msg":"sender: started","stream_id":"gc2qytpj"}
7
- {"time":"2025-03-28T20:38:44.910439284+08:00","level":"INFO","msg":"Starting system monitor"}
 
 
 
1
+ {"time":"2025-04-01T20:48:01.289721051+08:00","level":"INFO","msg":"stream: starting","core version":"0.19.8","symlink path":"../outputs/debug/wandb/run-20250401_204801-q89zozii/logs/debug-core.log"}
2
+ {"time":"2025-04-01T20:48:01.510627501+08:00","level":"INFO","msg":"created new stream","id":"q89zozii"}
3
+ {"time":"2025-04-01T20:48:01.510689964+08:00","level":"INFO","msg":"stream: started","id":"q89zozii"}
4
+ {"time":"2025-04-01T20:48:01.510740426+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"q89zozii"}
5
+ {"time":"2025-04-01T20:48:01.51076776+08:00","level":"INFO","msg":"handler: started","stream_id":"q89zozii"}
6
+ {"time":"2025-04-01T20:48:01.510769145+08:00","level":"INFO","msg":"sender: started","stream_id":"q89zozii"}
7
+ {"time":"2025-04-01T20:48:01.92178807+08:00","level":"INFO","msg":"Starting system monitor"}
8
+ {"time":"2025-04-03T06:57:27.965787009+08:00","level":"INFO","msg":"api: retrying HTTP error","status":502,"url":"https://api.wandb.ai/graphql","body":"\n<html><head>\n<meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n<title>502 Server Error</title>\n</head>\n<body text=#000000 bgcolor=#ffffff>\n<h1>Error: Server Error</h1>\n<h2>The server encountered a temporary error and could not complete your request.<p>Please try again in 30 seconds.</h2>\n<h2></h2>\n</body></html>\n"}
9
+ {"time":"2025-04-05T01:14:17.774572322+08:00","level":"INFO","msg":"api: retrying HTTP error","status":502,"url":"https://api.wandb.ai/files/htlou/align-anything/q89zozii/file_stream","body":"\n<html><head>\n<meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n<title>502 Server Error</title>\n</head>\n<body text=#000000 bgcolor=#ffffff>\n<h1>Error: Server Error</h1>\n<h2>The server encountered a temporary error and could not complete your request.<p>Please try again in 30 seconds.</h2>\n<h2></h2>\n</body></html>\n"}
debug.log CHANGED
@@ -1,26 +1,22 @@
1
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_setup.py:_flush():67] Current SDK version is 0.19.8
2
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_setup.py:_flush():67] Configure stats pid to 3871311
3
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_setup.py:_flush():67] Loading settings from /home/yangyaodong/.config/wandb/settings
4
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_setup.py:_flush():67] Loading settings from /aifs4su/yaodong/hantao/align-anything/scripts/wandb/settings
5
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_setup.py:_flush():67] Loading settings from environment variables
6
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_init.py:setup_run_log_directory():647] Logging user logs to ../outputs/test_13B/wandb/offline-run-20250328_203844-gc2qytpj/logs/debug.log
7
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_init.py:setup_run_log_directory():648] Logging internal logs to ../outputs/test_13B/wandb/offline-run-20250328_203844-gc2qytpj/logs/debug-internal.log
8
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_init.py:init():761] calling init triggers
9
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_init.py:init():766] wandb.init called with sweep_config: {}
10
- config: {'train_cfgs': {'save_checkpoint': True, 'load_checkpoint': False, 'ds_cfgs': 'ds_z3_config.json', 'epochs': 3, 'seed': 42, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 1, 'gradient_checkpointing': True, 'learning_rate': 1e-06, 'lr_scheduler_type': 'cosine', 'lr_warmup_ratio': 0.03, 'weight_decay': 0.0, 'adam_betas': [0.9, 0.95], 'bf16': True, 'fp16': False, 'eval_strategy': 'epoch', 'eval_interval': 10, 'regularization': 0.001, 'scale_coeff': 0.1, 'freeze_mm_proj': False, 'freeze_vision_tower': True, 'freeze_language_model': False}, 'data_cfgs': {'train_datasets': '/aifs4su/yaodong/hantao/datasets/AA_preference_vicuna-13b_cosi_cut/merged/top1-10', 'train_template': 'AA_TI2T_LLAVA', 'train_size': {}, 'train_split': 'train', 'train_name': 'text-image-to-text', 'train_data_files': {}, 'train_optional_args': [], 'eval_datasets': {}, 'eval_template': {}, 'eval_size': {}, 'eval_split': {}, 'eval_subset': {}, 'eval_data_files': {}, 'eval_optional_args': []}, 'logger_cfgs': {'log_type': 'wandb', 'log_project': 'align-anything', 'log_run_name': 'dpo', 'output_dir': '../outputs/test_13B', 'cache_dir': {}, 'save_total_limit': 6}, 'model_cfgs': {'model_name_or_path': '/aifs4su/yaodong/hantao/models/llava-v1.6-vicuna-13b-hf', 'trust_remote_code': True, 'model_max_length': 2048}, 'special_tokens': {}, '_wandb': {}}
11
- 2025-03-28 20:38:44,447 INFO MainThread:3871311 [wandb_init.py:init():784] starting backend
12
- 2025-03-28 20:38:44,657 INFO MainThread:3871311 [wandb_init.py:init():788] sending inform_init request
13
- 2025-03-28 20:38:44,660 INFO MainThread:3871311 [backend.py:_multiprocessing_setup():101] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
14
- 2025-03-28 20:38:44,660 INFO MainThread:3871311 [wandb_init.py:init():798] backend started and connected
15
- 2025-03-28 20:38:44,662 INFO MainThread:3871311 [wandb_init.py:init():891] updated telemetry
16
- 2025-03-28 20:38:44,672 INFO MainThread:3871311 [wandb_init.py:init():915] communicating run to backend with 90.0 second timeout
17
- 2025-03-28 20:38:44,908 INFO MainThread:3871311 [wandb_init.py:init():990] starting run threads in backend
18
- 2025-03-28 20:38:45,071 INFO MainThread:3871311 [wandb_run.py:_console_start():2375] atexit reg
19
- 2025-03-28 20:38:45,071 INFO MainThread:3871311 [wandb_run.py:_redirect():2227] redirect: wrap_raw
20
- 2025-03-28 20:38:45,071 INFO MainThread:3871311 [wandb_run.py:_redirect():2292] Wrapping output streams.
21
- 2025-03-28 20:38:45,071 INFO MainThread:3871311 [wandb_run.py:_redirect():2315] Redirects installed.
22
- 2025-03-28 20:38:45,074 INFO MainThread:3871311 [wandb_init.py:init():1032] run started, returning control to user process
23
- 2025-03-28 20:47:40,230 INFO MainThread:3871311 [wandb_run.py:_finish():2112] finishing run align-anything/gc2qytpj
24
- 2025-03-28 20:47:40,230 INFO MainThread:3871311 [wandb_run.py:_atexit_cleanup():2340] got exitcode: 0
25
- 2025-03-28 20:47:40,230 INFO MainThread:3871311 [wandb_run.py:_restore():2322] restore
26
- 2025-03-28 20:47:40,230 INFO MainThread:3871311 [wandb_run.py:_restore():2328] restore done
 
1
+ 2025-04-01 20:48:01,278 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Current SDK version is 0.19.8
2
+ 2025-04-01 20:48:01,278 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Configure stats pid to 2442497
3
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Loading settings from /home/yangyaodong/.config/wandb/settings
4
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Loading settings from /aifs4su/yaodong/hantao/align-anything/scripts/wandb/settings
5
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Loading settings from environment variables
6
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:setup_run_log_directory():647] Logging user logs to ../outputs/debug/wandb/run-20250401_204801-q89zozii/logs/debug.log
7
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:setup_run_log_directory():648] Logging internal logs to ../outputs/debug/wandb/run-20250401_204801-q89zozii/logs/debug-internal.log
8
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:init():761] calling init triggers
9
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:init():766] wandb.init called with sweep_config: {}
10
+ config: {'train_cfgs': {'save_checkpoint': False, 'load_checkpoint': False, 'ds_cfgs': 'ds_z3_config.json', 'epochs': 3000000, 'seed': 42, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 1, 'gradient_checkpointing': True, 'learning_rate': 1e-06, 'lr_scheduler_type': 'cosine', 'lr_warmup_ratio': 0.03, 'weight_decay': 0.0, 'adam_betas': [0.9, 0.95], 'bf16': True, 'fp16': False, 'eval_strategy': 'epoch', 'eval_interval': 10, 'regularization': 0.001, 'scale_coeff': 0.1, 'freeze_mm_proj': False, 'freeze_vision_tower': True, 'freeze_language_model': False}, 'data_cfgs': {'train_datasets': '/aifs4su/yaodong/hantao/datasets/AA_preference_vicuna-7b_cosi_cut/merged/top1-80', 'train_template': 'AA_TI2T_LLAVA', 'train_size': {}, 'train_split': 'train', 'train_name': 'text-image-to-text', 'train_data_files': {}, 'train_optional_args': [], 'eval_datasets': {}, 'eval_template': {}, 'eval_size': {}, 'eval_split': {}, 'eval_subset': {}, 'eval_data_files': {}, 'eval_optional_args': []}, 'logger_cfgs': {'log_type': 'wandb', 'log_project': 'align-anything', 'log_run_name': 'dpo', 'output_dir': '../outputs/debug', 'cache_dir': {}, 'save_total_limit': 3}, 'model_cfgs': {'model_name_or_path': '/aifs4su/yaodong/hantao/models/llava-v1.6-vicuna-7b-hf', 'trust_remote_code': True, 'model_max_length': 4096}, 'special_tokens': {}, '_wandb': {}}
11
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:init():784] starting backend
12
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:init():788] sending inform_init request
13
+ 2025-04-01 20:48:01,286 INFO MainThread:2442497 [backend.py:_multiprocessing_setup():101] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
14
+ 2025-04-01 20:48:01,286 INFO MainThread:2442497 [wandb_init.py:init():798] backend started and connected
15
+ 2025-04-01 20:48:01,293 INFO MainThread:2442497 [wandb_init.py:init():891] updated telemetry
16
+ 2025-04-01 20:48:01,316 INFO MainThread:2442497 [wandb_init.py:init():915] communicating run to backend with 90.0 second timeout
17
+ 2025-04-01 20:48:01,917 INFO MainThread:2442497 [wandb_init.py:init():990] starting run threads in backend
18
+ 2025-04-01 20:48:02,296 INFO MainThread:2442497 [wandb_run.py:_console_start():2375] atexit reg
19
+ 2025-04-01 20:48:02,296 INFO MainThread:2442497 [wandb_run.py:_redirect():2227] redirect: wrap_raw
20
+ 2025-04-01 20:48:02,296 INFO MainThread:2442497 [wandb_run.py:_redirect():2292] Wrapping output streams.
21
+ 2025-04-01 20:48:02,296 INFO MainThread:2442497 [wandb_run.py:_redirect():2315] Redirects installed.
22
+ 2025-04-01 20:48:02,302 INFO MainThread:2442497 [wandb_init.py:init():1032] run started, returning control to user process
 
 
 
 
run-20250401_170224-799gcj60/files/output.log ADDED
The diff for this file is too large to render. See raw diff
 
run-20250401_170224-799gcj60/files/requirements.txt ADDED
@@ -0,0 +1,258 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ maskrcnn_benchmark==0.0.0
2
+ webdataset==0.2.111
3
+ websockets==15.0.1
4
+ typer==0.15.2
5
+ blobfile==3.0.0
6
+ pooch==1.8.2
7
+ filelock==3.18.0
8
+ referencing==0.36.2
9
+ matplotlib==3.10.1
10
+ cachetools==5.5.2
11
+ python-dateutil==2.9.0.post0
12
+ gmpy2==2.2.1
13
+ uvloop==0.21.0
14
+ nvidia-cusparselt-cu12==0.6.2
15
+ clip==0.2.0
16
+ httpcore==1.0.7
17
+ charset-normalizer==3.3.2
18
+ torchlibrosa==0.1.0
19
+ contourpy==1.3.1
20
+ multiprocess==0.70.16
21
+ nest-asyncio==1.6.0
22
+ Werkzeug==3.1.3
23
+ aiofiles==23.2.1
24
+ six==1.17.0
25
+ torch==2.6.0
26
+ sse-starlette==2.2.1
27
+ typing_extensions==4.12.2
28
+ xgrammar==0.1.16
29
+ psutil==7.0.0
30
+ kiwisolver==1.4.8
31
+ moviepy==2.1.2
32
+ frozenlist==1.5.0
33
+ jiter==0.9.0
34
+ einops==0.8.1
35
+ flash_attn==2.7.4.post1
36
+ PySocks==1.7.1
37
+ regex==2024.11.6
38
+ markdown-it-py==3.0.0
39
+ ruff==0.11.2
40
+ docker-pycreds==0.4.0
41
+ nvidia-nvtx-cu12==12.4.127
42
+ pyparsing==3.2.3
43
+ resampy==0.4.3
44
+ tokenizers==0.21.0
45
+ frechet-audio-distance==0.1.2
46
+ aiohappyeyeballs==2.6.1
47
+ llamafactory==0.9.3.dev0
48
+ msgspec==0.19.0
49
+ httpx==0.28.1
50
+ encodec==0.1.1
51
+ ffmpy==0.5.0
52
+ jsonschema==4.23.0
53
+ imageio-ffmpeg==0.6.0
54
+ mkl_random==1.2.8
55
+ fairscale==0.4.13
56
+ soxr==0.5.0.post1
57
+ lark==1.2.2
58
+ gradio==5.21.0
59
+ absl-py==2.2.1
60
+ dnspython==2.7.0
61
+ networkx==3.4.2
62
+ h5py==3.13.0
63
+ hjson==3.1.0
64
+ tensorboard==2.19.0
65
+ aiosignal==1.3.2
66
+ pip==25.0
67
+ nvidia-cublas-cu12==12.4.5.8
68
+ llguidance==0.7.11
69
+ zipp==3.21.0
70
+ ftfy==6.3.1
71
+ peft==0.15.0
72
+ attrs==25.3.0
73
+ trl==0.9.6
74
+ requests==2.32.3
75
+ progressbar==2.5
76
+ sniffio==1.3.1
77
+ pycountry==24.6.1
78
+ lxml==5.3.1
79
+ starlette==0.46.1
80
+ pytest==7.2.0
81
+ Markdown==3.7
82
+ mdurl==0.1.2
83
+ pyzmq==26.3.0
84
+ safetensors==0.5.3
85
+ opencv-python==4.6.0.66
86
+ prometheus-fastapi-instrumentator==7.1.0
87
+ shellingham==1.5.4
88
+ torchvision==0.21.0
89
+ pluggy==1.5.0
90
+ timm==1.0.15
91
+ multidict==6.2.0
92
+ semantic-version==2.10.0
93
+ airportsdata==20250224
94
+ numba==0.60.0
95
+ MarkupSafe==2.1.5
96
+ pydantic_core==2.33.0
97
+ imageio==2.37.0
98
+ nvidia-nccl-cu12==2.21.5
99
+ dill==0.3.8
100
+ msgpack==1.1.0
101
+ sentry-sdk==2.24.1
102
+ rpds-py==0.24.0
103
+ grpcio==1.71.0
104
+ fastrlock==0.8.3
105
+ python-json-logger==3.3.0
106
+ cffi==1.17.1
107
+ gradio_client==1.7.2
108
+ PyYAML==6.0.2
109
+ tensorboard-data-server==0.7.2
110
+ termcolor==2.5.0
111
+ torchaudio==2.6.0
112
+ triton==3.2.0
113
+ fastapi==0.115.12
114
+ clint==0.5.1
115
+ lazy_loader==0.4
116
+ depyf==0.18.0
117
+ mkl_fft==1.3.11
118
+ annotated-types==0.7.0
119
+ scikit-learn==1.6.1
120
+ wget==3.2
121
+ setuptools==75.8.0
122
+ args==0.1.0
123
+ certifi==2025.1.31
124
+ click==8.1.8
125
+ python-dotenv==1.1.0
126
+ laion_clap==1.1.5
127
+ Pygments==2.19.1
128
+ tomlkit==0.13.2
129
+ idna==3.7
130
+ propcache==0.3.1
131
+ platformdirs==4.3.7
132
+ align-anything==0.0.1.dev0
133
+ ray==2.44.1
134
+ cloudpickle==3.1.1
135
+ deepspeed==0.16.5
136
+ smmap==5.0.2
137
+ distro==1.9.0
138
+ fonttools==4.56.0
139
+ typing-inspection==0.4.0
140
+ braceexpand==0.1.7
141
+ decorator==5.2.1
142
+ diskcache==5.6.3
143
+ yt-dlp==2025.3.27
144
+ shtab==1.7.1
145
+ gguf==0.10.0
146
+ interegular==0.3.3
147
+ compressed-tensors==0.9.2
148
+ pandas==2.2.3
149
+ huggingface-hub==0.29.3
150
+ pyarrow==19.0.1
151
+ lm-format-enforcer==0.10.11
152
+ GitPython==3.1.44
153
+ xxhash==3.5.0
154
+ packaging==24.2
155
+ setproctitle==1.3.5
156
+ llvmlite==0.43.0
157
+ tiktoken==0.9.0
158
+ mpmath==1.3.0
159
+ email_validator==2.2.0
160
+ nvidia-ml-py==12.570.86
161
+ pydantic==2.11.0
162
+ xformers==0.0.29.post2
163
+ httptools==0.6.4
164
+ librosa==0.11.0
165
+ pytorch-fid==0.3.0
166
+ hpsv2==1.2.0
167
+ nvidia-cudnn-cu12==9.1.0.70
168
+ iniconfig==2.1.0
169
+ sympy==1.13.1
170
+ safehttpx==0.1.6
171
+ jsonschema-specifications==2024.10.1
172
+ Jinja2==3.1.6
173
+ tyro==0.8.14
174
+ h11==0.14.0
175
+ aiohttp==3.11.14
176
+ diffusers==0.32.2
177
+ tqdm==4.67.1
178
+ blake3==1.0.4
179
+ vllm==0.8.2
180
+ scipy==1.10.1
181
+ audioread==3.0.1
182
+ proglog==0.1.10
183
+ fire==0.7.0
184
+ sentencepiece==0.2.0
185
+ pytz==2025.2
186
+ nvidia-nvjitlink-cu12==12.4.127
187
+ nvidia-curand-cu12==10.3.5.147
188
+ numpy==1.26.4
189
+ tzdata==2025.2
190
+ python-multipart==0.0.20
191
+ urllib3==2.3.0
192
+ pycryptodomex==3.22.0
193
+ yarl==1.18.3
194
+ outlines==0.1.11
195
+ nvidia-cusolver-cu12==11.6.1.9
196
+ pydub==0.25.1
197
+ mistral_common==1.5.4
198
+ pycparser==2.22
199
+ pytest-split==0.8.0
200
+ datasets==3.4.1
201
+ soundfile==0.13.1
202
+ transformers==4.50.0
203
+ image-reward==1.5
204
+ wcwidth==0.2.13
205
+ nvidia-cuda-nvrtc-cu12==12.4.127
206
+ groovy==0.1.2
207
+ nvidia-cuda-runtime-cu12==12.4.127
208
+ astor==0.8.1
209
+ anyio==4.9.0
210
+ wandb==0.19.8
211
+ joblib==1.4.2
212
+ fsspec==2024.12.0
213
+ accelerate==1.5.2
214
+ py-cpuinfo==9.0.0
215
+ docstring_parser==0.16
216
+ partial-json-parser==0.2.1.1.post5
217
+ nvidia-cusparse-cu12==12.3.1.170
218
+ protobuf==3.20.3
219
+ outlines_core==0.1.26
220
+ nvidia-cufft-cu12==11.2.1.3
221
+ cycler==0.12.1
222
+ uvicorn==0.34.0
223
+ orjson==3.10.16
224
+ av==14.2.0
225
+ Brotli==1.0.9
226
+ cupy-cuda12x==13.4.1
227
+ openai==1.69.0
228
+ rich==13.9.4
229
+ importlib_metadata==8.6.1
230
+ ninja==1.11.1.4
231
+ wheel==0.45.1
232
+ pillow==10.4.0
233
+ prometheus_client==0.21.1
234
+ threadpoolctl==3.6.0
235
+ gitdb==4.0.12
236
+ watchfiles==1.0.4
237
+ nvidia-cuda-cupti-cu12==12.4.127
238
+ opencv-python-headless==4.11.0.86
239
+ mkl-service==2.4.0
240
+ rich-toolkit==0.14.0
241
+ fastapi-cli==0.0.7
242
+ llamafactory==0.9.3.dev0
243
+ typing_extensions==4.12.2
244
+ tomli==2.0.1
245
+ zipp==3.19.2
246
+ wheel==0.43.0
247
+ jaraco.text==3.12.1
248
+ packaging==24.2
249
+ autocommand==2.2.2
250
+ jaraco.functools==4.0.1
251
+ jaraco.collections==5.1.0
252
+ platformdirs==4.2.2
253
+ more-itertools==10.3.0
254
+ inflect==7.3.1
255
+ jaraco.context==5.3.0
256
+ typeguard==4.3.0
257
+ backports.tarfile==1.2.0
258
+ importlib_metadata==8.0.0
run-20250401_170224-799gcj60/files/wandb-metadata.json ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "os": "Linux-5.15.0-1040-nvidia-x86_64-with-glibc2.35",
3
+ "python": "CPython 3.11.11",
4
+ "startedAt": "2025-04-01T09:02:24.644727Z",
5
+ "args": [
6
+ "--local_rank=0",
7
+ "--model_name_or_path",
8
+ "/aifs4su/yaodong/hantao/models/llava-v1.6-vicuna-7b-hf",
9
+ "--train_datasets",
10
+ "/aifs4su/yaodong/hantao/datasets/AA_preference_vicuna-7b_cosi_cut/merged/top1-80",
11
+ "--train_template",
12
+ "AA_TI2T_LLAVA",
13
+ "--train_split",
14
+ "train",
15
+ "--train_name",
16
+ "text-image-to-text",
17
+ "--output_dir",
18
+ "../outputs/debug",
19
+ "--save_total_limit",
20
+ "3",
21
+ "--train_batch_size",
22
+ "8",
23
+ "--epochs",
24
+ "3000000"
25
+ ],
26
+ "program": "-m align_anything.trainers.text_image_to_text.dpo",
27
+ "git": {
28
+ "remote": "git@github.com-hantao:PKU-Alignment/align-anything.git",
29
+ "commit": "106588f9802757a3283c1aff1f33ea9afd737f31"
30
+ },
31
+ "email": "2200017789@stu.pku.edu.cn",
32
+ "root": "../outputs/debug",
33
+ "host": "dgx-092",
34
+ "executable": "/aifs4su/yaodong/miniconda3/envs/hantao_llama/bin/python",
35
+ "cpu_count": 112,
36
+ "cpu_count_logical": 224,
37
+ "gpu": "NVIDIA H800",
38
+ "gpu_count": 8,
39
+ "disk": {
40
+ "/": {
41
+ "total": "1888556142592",
42
+ "used": "149111373824"
43
+ }
44
+ },
45
+ "memory": {
46
+ "total": "2164195454976"
47
+ },
48
+ "cpu": {
49
+ "count": 112,
50
+ "countLogical": 224
51
+ },
52
+ "gpu_nvidia": [
53
+ {
54
+ "name": "NVIDIA H800",
55
+ "memoryTotal": "85520809984",
56
+ "cudaCores": 16896,
57
+ "architecture": "Hopper"
58
+ },
59
+ {
60
+ "name": "NVIDIA H800",
61
+ "memoryTotal": "85520809984",
62
+ "cudaCores": 16896,
63
+ "architecture": "Hopper"
64
+ },
65
+ {
66
+ "name": "NVIDIA H800",
67
+ "memoryTotal": "85520809984",
68
+ "cudaCores": 16896,
69
+ "architecture": "Hopper"
70
+ },
71
+ {
72
+ "name": "NVIDIA H800",
73
+ "memoryTotal": "85520809984",
74
+ "cudaCores": 16896,
75
+ "architecture": "Hopper"
76
+ },
77
+ {
78
+ "name": "NVIDIA H800",
79
+ "memoryTotal": "85520809984",
80
+ "cudaCores": 16896,
81
+ "architecture": "Hopper"
82
+ },
83
+ {
84
+ "name": "NVIDIA H800",
85
+ "memoryTotal": "85520809984",
86
+ "cudaCores": 16896,
87
+ "architecture": "Hopper"
88
+ },
89
+ {
90
+ "name": "NVIDIA H800",
91
+ "memoryTotal": "85520809984",
92
+ "cudaCores": 16896,
93
+ "architecture": "Hopper"
94
+ },
95
+ {
96
+ "name": "NVIDIA H800",
97
+ "memoryTotal": "85520809984",
98
+ "cudaCores": 16896,
99
+ "architecture": "Hopper"
100
+ }
101
+ ],
102
+ "slurm": {
103
+ "conf": "/cm/shared/apps/slurm/var/etc/slurm/slurm.conf"
104
+ },
105
+ "cudaVersion": "12.2"
106
+ }
run-20250401_170224-799gcj60/logs/debug-core.log ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {"time":"2025-04-01T17:02:24.03376922+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpd1qy9mk6/port-351970.txt","pid":351970,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false}
2
+ {"time":"2025-04-01T17:02:24.034941698+08:00","level":"INFO","msg":"Will exit if parent process dies.","ppid":351970}
3
+ {"time":"2025-04-01T17:02:24.034930618+08:00","level":"INFO","msg":"server is running","addr":{"IP":"127.0.0.1","Port":35239,"Zone":""}}
4
+ {"time":"2025-04-01T17:02:24.191444247+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"127.0.0.1:48758"}
5
+ {"time":"2025-04-01T17:02:24.645799183+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"799gcj60","id":"127.0.0.1:48758"}
6
+ {"time":"2025-04-01T17:02:24.867535803+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"799gcj60","id":"127.0.0.1:48758"}
7
+ {"time":"2025-04-01T19:21:15.95647246+08:00","level":"INFO","msg":"received shutdown signal","signal":15}
run-20250401_170224-799gcj60/logs/debug-internal.log ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {"time":"2025-04-01T17:02:24.647157427+08:00","level":"INFO","msg":"stream: starting","core version":"0.19.8","symlink path":"../outputs/debug/wandb/run-20250401_170224-799gcj60/logs/debug-core.log"}
2
+ {"time":"2025-04-01T17:02:24.867468348+08:00","level":"INFO","msg":"created new stream","id":"799gcj60"}
3
+ {"time":"2025-04-01T17:02:24.867527171+08:00","level":"INFO","msg":"stream: started","id":"799gcj60"}
4
+ {"time":"2025-04-01T17:02:24.867549376+08:00","level":"INFO","msg":"sender: started","stream_id":"799gcj60"}
5
+ {"time":"2025-04-01T17:02:24.867556659+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"799gcj60"}
6
+ {"time":"2025-04-01T17:02:24.867574339+08:00","level":"INFO","msg":"handler: started","stream_id":"799gcj60"}
7
+ {"time":"2025-04-01T17:02:25.2785929+08:00","level":"INFO","msg":"Starting system monitor"}
run-20250401_170224-799gcj60/logs/debug.log ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2025-04-01 17:02:24,633 INFO MainThread:351970 [wandb_setup.py:_flush():67] Current SDK version is 0.19.8
2
+ 2025-04-01 17:02:24,633 INFO MainThread:351970 [wandb_setup.py:_flush():67] Configure stats pid to 351970
3
+ 2025-04-01 17:02:24,633 INFO MainThread:351970 [wandb_setup.py:_flush():67] Loading settings from /home/yangyaodong/.config/wandb/settings
4
+ 2025-04-01 17:02:24,634 INFO MainThread:351970 [wandb_setup.py:_flush():67] Loading settings from /aifs4su/yaodong/hantao/align-anything/scripts/wandb/settings
5
+ 2025-04-01 17:02:24,634 INFO MainThread:351970 [wandb_setup.py:_flush():67] Loading settings from environment variables
6
+ 2025-04-01 17:02:24,634 INFO MainThread:351970 [wandb_init.py:setup_run_log_directory():647] Logging user logs to ../outputs/debug/wandb/run-20250401_170224-799gcj60/logs/debug.log
7
+ 2025-04-01 17:02:24,634 INFO MainThread:351970 [wandb_init.py:setup_run_log_directory():648] Logging internal logs to ../outputs/debug/wandb/run-20250401_170224-799gcj60/logs/debug-internal.log
8
+ 2025-04-01 17:02:24,634 INFO MainThread:351970 [wandb_init.py:init():761] calling init triggers
9
+ 2025-04-01 17:02:24,634 INFO MainThread:351970 [wandb_init.py:init():766] wandb.init called with sweep_config: {}
10
+ config: {'train_cfgs': {'save_checkpoint': False, 'load_checkpoint': False, 'ds_cfgs': 'ds_z3_config.json', 'epochs': 3000000, 'seed': 42, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 1, 'gradient_checkpointing': True, 'learning_rate': 1e-06, 'lr_scheduler_type': 'cosine', 'lr_warmup_ratio': 0.03, 'weight_decay': 0.0, 'adam_betas': [0.9, 0.95], 'bf16': True, 'fp16': False, 'eval_strategy': 'epoch', 'eval_interval': 10, 'regularization': 0.001, 'scale_coeff': 0.1, 'freeze_mm_proj': False, 'freeze_vision_tower': True, 'freeze_language_model': False}, 'data_cfgs': {'train_datasets': '/aifs4su/yaodong/hantao/datasets/AA_preference_vicuna-7b_cosi_cut/merged/top1-80', 'train_template': 'AA_TI2T_LLAVA', 'train_size': {}, 'train_split': 'train', 'train_name': 'text-image-to-text', 'train_data_files': {}, 'train_optional_args': [], 'eval_datasets': {}, 'eval_template': {}, 'eval_size': {}, 'eval_split': {}, 'eval_subset': {}, 'eval_data_files': {}, 'eval_optional_args': []}, 'logger_cfgs': {'log_type': 'wandb', 'log_project': 'align-anything', 'log_run_name': 'dpo', 'output_dir': '../outputs/debug', 'cache_dir': {}, 'save_total_limit': 3}, 'model_cfgs': {'model_name_or_path': '/aifs4su/yaodong/hantao/models/llava-v1.6-vicuna-7b-hf', 'trust_remote_code': True, 'model_max_length': 4096}, 'special_tokens': {}, '_wandb': {}}
11
+ 2025-04-01 17:02:24,634 INFO MainThread:351970 [wandb_init.py:init():784] starting backend
12
+ 2025-04-01 17:02:24,634 INFO MainThread:351970 [wandb_init.py:init():788] sending inform_init request
13
+ 2025-04-01 17:02:24,644 INFO MainThread:351970 [backend.py:_multiprocessing_setup():101] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
14
+ 2025-04-01 17:02:24,644 INFO MainThread:351970 [wandb_init.py:init():798] backend started and connected
15
+ 2025-04-01 17:02:24,646 INFO MainThread:351970 [wandb_init.py:init():891] updated telemetry
16
+ 2025-04-01 17:02:24,657 INFO MainThread:351970 [wandb_init.py:init():915] communicating run to backend with 90.0 second timeout
17
+ 2025-04-01 17:02:25,272 INFO MainThread:351970 [wandb_init.py:init():990] starting run threads in backend
18
+ 2025-04-01 17:02:25,685 INFO MainThread:351970 [wandb_run.py:_console_start():2375] atexit reg
19
+ 2025-04-01 17:02:25,686 INFO MainThread:351970 [wandb_run.py:_redirect():2227] redirect: wrap_raw
20
+ 2025-04-01 17:02:25,687 INFO MainThread:351970 [wandb_run.py:_redirect():2292] Wrapping output streams.
21
+ 2025-04-01 17:02:25,687 INFO MainThread:351970 [wandb_run.py:_redirect():2315] Redirects installed.
22
+ 2025-04-01 17:02:25,696 INFO MainThread:351970 [wandb_init.py:init():1032] run started, returning control to user process
run-20250401_170224-799gcj60/run-799gcj60.wandb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00fdbacfbb075046e4e5c31ffbb7cbac0d413ad4426070e0f5758bf5d40d35ac
3
+ size 7307264
run-20250401_204801-q89zozii/files/output.log ADDED
The diff for this file is too large to render. See raw diff
 
run-20250401_204801-q89zozii/files/requirements.txt ADDED
@@ -0,0 +1,258 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ maskrcnn_benchmark==0.0.0
2
+ webdataset==0.2.111
3
+ websockets==15.0.1
4
+ typer==0.15.2
5
+ blobfile==3.0.0
6
+ pooch==1.8.2
7
+ filelock==3.18.0
8
+ referencing==0.36.2
9
+ matplotlib==3.10.1
10
+ cachetools==5.5.2
11
+ python-dateutil==2.9.0.post0
12
+ gmpy2==2.2.1
13
+ uvloop==0.21.0
14
+ nvidia-cusparselt-cu12==0.6.2
15
+ clip==0.2.0
16
+ httpcore==1.0.7
17
+ charset-normalizer==3.3.2
18
+ torchlibrosa==0.1.0
19
+ contourpy==1.3.1
20
+ multiprocess==0.70.16
21
+ nest-asyncio==1.6.0
22
+ Werkzeug==3.1.3
23
+ aiofiles==23.2.1
24
+ six==1.17.0
25
+ torch==2.6.0
26
+ sse-starlette==2.2.1
27
+ typing_extensions==4.12.2
28
+ xgrammar==0.1.16
29
+ psutil==7.0.0
30
+ kiwisolver==1.4.8
31
+ moviepy==2.1.2
32
+ frozenlist==1.5.0
33
+ jiter==0.9.0
34
+ einops==0.8.1
35
+ flash_attn==2.7.4.post1
36
+ PySocks==1.7.1
37
+ regex==2024.11.6
38
+ markdown-it-py==3.0.0
39
+ ruff==0.11.2
40
+ docker-pycreds==0.4.0
41
+ nvidia-nvtx-cu12==12.4.127
42
+ pyparsing==3.2.3
43
+ resampy==0.4.3
44
+ tokenizers==0.21.0
45
+ frechet-audio-distance==0.1.2
46
+ aiohappyeyeballs==2.6.1
47
+ llamafactory==0.9.3.dev0
48
+ msgspec==0.19.0
49
+ httpx==0.28.1
50
+ encodec==0.1.1
51
+ ffmpy==0.5.0
52
+ jsonschema==4.23.0
53
+ imageio-ffmpeg==0.6.0
54
+ mkl_random==1.2.8
55
+ fairscale==0.4.13
56
+ soxr==0.5.0.post1
57
+ lark==1.2.2
58
+ gradio==5.21.0
59
+ absl-py==2.2.1
60
+ dnspython==2.7.0
61
+ networkx==3.4.2
62
+ h5py==3.13.0
63
+ hjson==3.1.0
64
+ tensorboard==2.19.0
65
+ aiosignal==1.3.2
66
+ pip==25.0
67
+ nvidia-cublas-cu12==12.4.5.8
68
+ llguidance==0.7.11
69
+ zipp==3.21.0
70
+ ftfy==6.3.1
71
+ peft==0.15.0
72
+ attrs==25.3.0
73
+ trl==0.9.6
74
+ requests==2.32.3
75
+ progressbar==2.5
76
+ sniffio==1.3.1
77
+ pycountry==24.6.1
78
+ lxml==5.3.1
79
+ starlette==0.46.1
80
+ pytest==7.2.0
81
+ Markdown==3.7
82
+ mdurl==0.1.2
83
+ pyzmq==26.3.0
84
+ safetensors==0.5.3
85
+ opencv-python==4.6.0.66
86
+ prometheus-fastapi-instrumentator==7.1.0
87
+ shellingham==1.5.4
88
+ torchvision==0.21.0
89
+ pluggy==1.5.0
90
+ timm==1.0.15
91
+ multidict==6.2.0
92
+ semantic-version==2.10.0
93
+ airportsdata==20250224
94
+ numba==0.60.0
95
+ MarkupSafe==2.1.5
96
+ pydantic_core==2.33.0
97
+ imageio==2.37.0
98
+ nvidia-nccl-cu12==2.21.5
99
+ dill==0.3.8
100
+ msgpack==1.1.0
101
+ sentry-sdk==2.24.1
102
+ rpds-py==0.24.0
103
+ grpcio==1.71.0
104
+ fastrlock==0.8.3
105
+ python-json-logger==3.3.0
106
+ cffi==1.17.1
107
+ gradio_client==1.7.2
108
+ PyYAML==6.0.2
109
+ tensorboard-data-server==0.7.2
110
+ termcolor==2.5.0
111
+ torchaudio==2.6.0
112
+ triton==3.2.0
113
+ fastapi==0.115.12
114
+ clint==0.5.1
115
+ lazy_loader==0.4
116
+ depyf==0.18.0
117
+ mkl_fft==1.3.11
118
+ annotated-types==0.7.0
119
+ scikit-learn==1.6.1
120
+ wget==3.2
121
+ setuptools==75.8.0
122
+ args==0.1.0
123
+ certifi==2025.1.31
124
+ click==8.1.8
125
+ python-dotenv==1.1.0
126
+ laion_clap==1.1.5
127
+ Pygments==2.19.1
128
+ tomlkit==0.13.2
129
+ idna==3.7
130
+ propcache==0.3.1
131
+ platformdirs==4.3.7
132
+ align-anything==0.0.1.dev0
133
+ ray==2.44.1
134
+ cloudpickle==3.1.1
135
+ deepspeed==0.16.5
136
+ smmap==5.0.2
137
+ distro==1.9.0
138
+ fonttools==4.56.0
139
+ typing-inspection==0.4.0
140
+ braceexpand==0.1.7
141
+ decorator==5.2.1
142
+ diskcache==5.6.3
143
+ yt-dlp==2025.3.27
144
+ shtab==1.7.1
145
+ gguf==0.10.0
146
+ interegular==0.3.3
147
+ compressed-tensors==0.9.2
148
+ pandas==2.2.3
149
+ huggingface-hub==0.29.3
150
+ pyarrow==19.0.1
151
+ lm-format-enforcer==0.10.11
152
+ GitPython==3.1.44
153
+ xxhash==3.5.0
154
+ packaging==24.2
155
+ setproctitle==1.3.5
156
+ llvmlite==0.43.0
157
+ tiktoken==0.9.0
158
+ mpmath==1.3.0
159
+ email_validator==2.2.0
160
+ nvidia-ml-py==12.570.86
161
+ pydantic==2.11.0
162
+ xformers==0.0.29.post2
163
+ httptools==0.6.4
164
+ librosa==0.11.0
165
+ pytorch-fid==0.3.0
166
+ hpsv2==1.2.0
167
+ nvidia-cudnn-cu12==9.1.0.70
168
+ iniconfig==2.1.0
169
+ sympy==1.13.1
170
+ safehttpx==0.1.6
171
+ jsonschema-specifications==2024.10.1
172
+ Jinja2==3.1.6
173
+ tyro==0.8.14
174
+ h11==0.14.0
175
+ aiohttp==3.11.14
176
+ diffusers==0.32.2
177
+ tqdm==4.67.1
178
+ blake3==1.0.4
179
+ vllm==0.8.2
180
+ scipy==1.10.1
181
+ audioread==3.0.1
182
+ proglog==0.1.10
183
+ fire==0.7.0
184
+ sentencepiece==0.2.0
185
+ pytz==2025.2
186
+ nvidia-nvjitlink-cu12==12.4.127
187
+ nvidia-curand-cu12==10.3.5.147
188
+ numpy==1.26.4
189
+ tzdata==2025.2
190
+ python-multipart==0.0.20
191
+ urllib3==2.3.0
192
+ pycryptodomex==3.22.0
193
+ yarl==1.18.3
194
+ outlines==0.1.11
195
+ nvidia-cusolver-cu12==11.6.1.9
196
+ pydub==0.25.1
197
+ mistral_common==1.5.4
198
+ pycparser==2.22
199
+ pytest-split==0.8.0
200
+ datasets==3.4.1
201
+ soundfile==0.13.1
202
+ transformers==4.50.0
203
+ image-reward==1.5
204
+ wcwidth==0.2.13
205
+ nvidia-cuda-nvrtc-cu12==12.4.127
206
+ groovy==0.1.2
207
+ nvidia-cuda-runtime-cu12==12.4.127
208
+ astor==0.8.1
209
+ anyio==4.9.0
210
+ wandb==0.19.8
211
+ joblib==1.4.2
212
+ fsspec==2024.12.0
213
+ accelerate==1.5.2
214
+ py-cpuinfo==9.0.0
215
+ docstring_parser==0.16
216
+ partial-json-parser==0.2.1.1.post5
217
+ nvidia-cusparse-cu12==12.3.1.170
218
+ protobuf==3.20.3
219
+ outlines_core==0.1.26
220
+ nvidia-cufft-cu12==11.2.1.3
221
+ cycler==0.12.1
222
+ uvicorn==0.34.0
223
+ orjson==3.10.16
224
+ av==14.2.0
225
+ Brotli==1.0.9
226
+ cupy-cuda12x==13.4.1
227
+ openai==1.69.0
228
+ rich==13.9.4
229
+ importlib_metadata==8.6.1
230
+ ninja==1.11.1.4
231
+ wheel==0.45.1
232
+ pillow==10.4.0
233
+ prometheus_client==0.21.1
234
+ threadpoolctl==3.6.0
235
+ gitdb==4.0.12
236
+ watchfiles==1.0.4
237
+ nvidia-cuda-cupti-cu12==12.4.127
238
+ opencv-python-headless==4.11.0.86
239
+ mkl-service==2.4.0
240
+ rich-toolkit==0.14.0
241
+ fastapi-cli==0.0.7
242
+ llamafactory==0.9.3.dev0
243
+ typing_extensions==4.12.2
244
+ tomli==2.0.1
245
+ zipp==3.19.2
246
+ wheel==0.43.0
247
+ jaraco.text==3.12.1
248
+ packaging==24.2
249
+ autocommand==2.2.2
250
+ jaraco.functools==4.0.1
251
+ jaraco.collections==5.1.0
252
+ platformdirs==4.2.2
253
+ more-itertools==10.3.0
254
+ inflect==7.3.1
255
+ jaraco.context==5.3.0
256
+ typeguard==4.3.0
257
+ backports.tarfile==1.2.0
258
+ importlib_metadata==8.0.0
run-20250401_204801-q89zozii/files/wandb-metadata.json ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "os": "Linux-5.15.0-1040-nvidia-x86_64-with-glibc2.35",
3
+ "python": "CPython 3.11.11",
4
+ "startedAt": "2025-04-01T12:48:01.287248Z",
5
+ "args": [
6
+ "--local_rank=0",
7
+ "--model_name_or_path",
8
+ "/aifs4su/yaodong/hantao/models/llava-v1.6-vicuna-7b-hf",
9
+ "--train_datasets",
10
+ "/aifs4su/yaodong/hantao/datasets/AA_preference_vicuna-7b_cosi_cut/merged/top1-80",
11
+ "--train_template",
12
+ "AA_TI2T_LLAVA",
13
+ "--train_split",
14
+ "train",
15
+ "--train_name",
16
+ "text-image-to-text",
17
+ "--output_dir",
18
+ "../outputs/debug",
19
+ "--save_total_limit",
20
+ "3",
21
+ "--train_batch_size",
22
+ "8",
23
+ "--epochs",
24
+ "3000000"
25
+ ],
26
+ "program": "-m align_anything.trainers.text_image_to_text.dpo",
27
+ "git": {
28
+ "remote": "git@github.com-hantao:PKU-Alignment/align-anything.git",
29
+ "commit": "106588f9802757a3283c1aff1f33ea9afd737f31"
30
+ },
31
+ "email": "2200017789@stu.pku.edu.cn",
32
+ "root": "../outputs/debug",
33
+ "host": "dgx-093",
34
+ "executable": "/aifs4su/yaodong/miniconda3/envs/hantao_llama/bin/python",
35
+ "cpu_count": 112,
36
+ "cpu_count_logical": 224,
37
+ "gpu": "NVIDIA H800",
38
+ "gpu_count": 8,
39
+ "disk": {
40
+ "/": {
41
+ "total": "1888556142592",
42
+ "used": "16262791168"
43
+ }
44
+ },
45
+ "memory": {
46
+ "total": "2164195565568"
47
+ },
48
+ "cpu": {
49
+ "count": 112,
50
+ "countLogical": 224
51
+ },
52
+ "gpu_nvidia": [
53
+ {
54
+ "name": "NVIDIA H800",
55
+ "memoryTotal": "85520809984",
56
+ "cudaCores": 16896,
57
+ "architecture": "Hopper"
58
+ },
59
+ {
60
+ "name": "NVIDIA H800",
61
+ "memoryTotal": "85520809984",
62
+ "cudaCores": 16896,
63
+ "architecture": "Hopper"
64
+ },
65
+ {
66
+ "name": "NVIDIA H800",
67
+ "memoryTotal": "85520809984",
68
+ "cudaCores": 16896,
69
+ "architecture": "Hopper"
70
+ },
71
+ {
72
+ "name": "NVIDIA H800",
73
+ "memoryTotal": "85520809984",
74
+ "cudaCores": 16896,
75
+ "architecture": "Hopper"
76
+ },
77
+ {
78
+ "name": "NVIDIA H800",
79
+ "memoryTotal": "85520809984",
80
+ "cudaCores": 16896,
81
+ "architecture": "Hopper"
82
+ },
83
+ {
84
+ "name": "NVIDIA H800",
85
+ "memoryTotal": "85520809984",
86
+ "cudaCores": 16896,
87
+ "architecture": "Hopper"
88
+ },
89
+ {
90
+ "name": "NVIDIA H800",
91
+ "memoryTotal": "85520809984",
92
+ "cudaCores": 16896,
93
+ "architecture": "Hopper"
94
+ },
95
+ {
96
+ "name": "NVIDIA H800",
97
+ "memoryTotal": "85520809984",
98
+ "cudaCores": 16896,
99
+ "architecture": "Hopper"
100
+ }
101
+ ],
102
+ "slurm": {
103
+ "conf": "/cm/shared/apps/slurm/var/etc/slurm/slurm.conf"
104
+ },
105
+ "cudaVersion": "12.2"
106
+ }
run-20250401_204801-q89zozii/logs/debug-core.log ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {"time":"2025-04-01T20:48:00.622799986+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpv_eb2tse/port-2442497.txt","pid":2442497,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false}
2
+ {"time":"2025-04-01T20:48:00.623998008+08:00","level":"INFO","msg":"Will exit if parent process dies.","ppid":2442497}
3
+ {"time":"2025-04-01T20:48:00.6239971+08:00","level":"INFO","msg":"server is running","addr":{"IP":"127.0.0.1","Port":42943,"Zone":""}}
4
+ {"time":"2025-04-01T20:48:00.779401958+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"127.0.0.1:61196"}
5
+ {"time":"2025-04-01T20:48:01.288492776+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"q89zozii","id":"127.0.0.1:61196"}
6
+ {"time":"2025-04-01T20:48:01.510696441+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"q89zozii","id":"127.0.0.1:61196"}
7
+ {"time":"2025-04-05T10:09:02.334506919+08:00","level":"INFO","msg":"received shutdown signal","signal":15}
run-20250401_204801-q89zozii/logs/debug-internal.log ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {"time":"2025-04-01T20:48:01.289721051+08:00","level":"INFO","msg":"stream: starting","core version":"0.19.8","symlink path":"../outputs/debug/wandb/run-20250401_204801-q89zozii/logs/debug-core.log"}
2
+ {"time":"2025-04-01T20:48:01.510627501+08:00","level":"INFO","msg":"created new stream","id":"q89zozii"}
3
+ {"time":"2025-04-01T20:48:01.510689964+08:00","level":"INFO","msg":"stream: started","id":"q89zozii"}
4
+ {"time":"2025-04-01T20:48:01.510740426+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"q89zozii"}
5
+ {"time":"2025-04-01T20:48:01.51076776+08:00","level":"INFO","msg":"handler: started","stream_id":"q89zozii"}
6
+ {"time":"2025-04-01T20:48:01.510769145+08:00","level":"INFO","msg":"sender: started","stream_id":"q89zozii"}
7
+ {"time":"2025-04-01T20:48:01.92178807+08:00","level":"INFO","msg":"Starting system monitor"}
8
+ {"time":"2025-04-03T06:57:27.965787009+08:00","level":"INFO","msg":"api: retrying HTTP error","status":502,"url":"https://api.wandb.ai/graphql","body":"\n<html><head>\n<meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n<title>502 Server Error</title>\n</head>\n<body text=#000000 bgcolor=#ffffff>\n<h1>Error: Server Error</h1>\n<h2>The server encountered a temporary error and could not complete your request.<p>Please try again in 30 seconds.</h2>\n<h2></h2>\n</body></html>\n"}
9
+ {"time":"2025-04-05T01:14:17.774572322+08:00","level":"INFO","msg":"api: retrying HTTP error","status":502,"url":"https://api.wandb.ai/files/htlou/align-anything/q89zozii/file_stream","body":"\n<html><head>\n<meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n<title>502 Server Error</title>\n</head>\n<body text=#000000 bgcolor=#ffffff>\n<h1>Error: Server Error</h1>\n<h2>The server encountered a temporary error and could not complete your request.<p>Please try again in 30 seconds.</h2>\n<h2></h2>\n</body></html>\n"}
run-20250401_204801-q89zozii/logs/debug.log ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2025-04-01 20:48:01,278 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Current SDK version is 0.19.8
2
+ 2025-04-01 20:48:01,278 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Configure stats pid to 2442497
3
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Loading settings from /home/yangyaodong/.config/wandb/settings
4
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Loading settings from /aifs4su/yaodong/hantao/align-anything/scripts/wandb/settings
5
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_setup.py:_flush():67] Loading settings from environment variables
6
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:setup_run_log_directory():647] Logging user logs to ../outputs/debug/wandb/run-20250401_204801-q89zozii/logs/debug.log
7
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:setup_run_log_directory():648] Logging internal logs to ../outputs/debug/wandb/run-20250401_204801-q89zozii/logs/debug-internal.log
8
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:init():761] calling init triggers
9
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:init():766] wandb.init called with sweep_config: {}
10
+ config: {'train_cfgs': {'save_checkpoint': False, 'load_checkpoint': False, 'ds_cfgs': 'ds_z3_config.json', 'epochs': 3000000, 'seed': 42, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 1, 'gradient_checkpointing': True, 'learning_rate': 1e-06, 'lr_scheduler_type': 'cosine', 'lr_warmup_ratio': 0.03, 'weight_decay': 0.0, 'adam_betas': [0.9, 0.95], 'bf16': True, 'fp16': False, 'eval_strategy': 'epoch', 'eval_interval': 10, 'regularization': 0.001, 'scale_coeff': 0.1, 'freeze_mm_proj': False, 'freeze_vision_tower': True, 'freeze_language_model': False}, 'data_cfgs': {'train_datasets': '/aifs4su/yaodong/hantao/datasets/AA_preference_vicuna-7b_cosi_cut/merged/top1-80', 'train_template': 'AA_TI2T_LLAVA', 'train_size': {}, 'train_split': 'train', 'train_name': 'text-image-to-text', 'train_data_files': {}, 'train_optional_args': [], 'eval_datasets': {}, 'eval_template': {}, 'eval_size': {}, 'eval_split': {}, 'eval_subset': {}, 'eval_data_files': {}, 'eval_optional_args': []}, 'logger_cfgs': {'log_type': 'wandb', 'log_project': 'align-anything', 'log_run_name': 'dpo', 'output_dir': '../outputs/debug', 'cache_dir': {}, 'save_total_limit': 3}, 'model_cfgs': {'model_name_or_path': '/aifs4su/yaodong/hantao/models/llava-v1.6-vicuna-7b-hf', 'trust_remote_code': True, 'model_max_length': 4096}, 'special_tokens': {}, '_wandb': {}}
11
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:init():784] starting backend
12
+ 2025-04-01 20:48:01,279 INFO MainThread:2442497 [wandb_init.py:init():788] sending inform_init request
13
+ 2025-04-01 20:48:01,286 INFO MainThread:2442497 [backend.py:_multiprocessing_setup():101] multiprocessing start_methods=fork,spawn,forkserver, using: spawn
14
+ 2025-04-01 20:48:01,286 INFO MainThread:2442497 [wandb_init.py:init():798] backend started and connected
15
+ 2025-04-01 20:48:01,293 INFO MainThread:2442497 [wandb_init.py:init():891] updated telemetry
16
+ 2025-04-01 20:48:01,316 INFO MainThread:2442497 [wandb_init.py:init():915] communicating run to backend with 90.0 second timeout
17
+ 2025-04-01 20:48:01,917 INFO MainThread:2442497 [wandb_init.py:init():990] starting run threads in backend
18
+ 2025-04-01 20:48:02,296 INFO MainThread:2442497 [wandb_run.py:_console_start():2375] atexit reg
19
+ 2025-04-01 20:48:02,296 INFO MainThread:2442497 [wandb_run.py:_redirect():2227] redirect: wrap_raw
20
+ 2025-04-01 20:48:02,296 INFO MainThread:2442497 [wandb_run.py:_redirect():2292] Wrapping output streams.
21
+ 2025-04-01 20:48:02,296 INFO MainThread:2442497 [wandb_run.py:_redirect():2315] Redirects installed.
22
+ 2025-04-01 20:48:02,302 INFO MainThread:2442497 [wandb_init.py:init():1032] run started, returning control to user process
run-20250401_204801-q89zozii/run-q89zozii.wandb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01c9ea7a997d97fc2ab3f0cdcab0706197b47a4a8ebf9754757837856a15dc0a
3
+ size 277643264