AbhishekGhosh1998 commited on
Commit
3e531b2
·
verified ·
1 Parent(s): a19ea07

granularity results granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace @ 2026-09-21T09:11:34Z from sp-aiinfra-interactive-aghosh-8-mb99k-8xb200-moe-cg-0

Browse files
Files changed (14) hide show
  1. .gitattributes +1 -0
  2. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/manifest.txt +19 -0
  3. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/pip_freeze.txt +317 -0
  4. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/provenance.txt +90 -0
  5. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.nsys-rep +3 -0
  6. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/moonep_async_overlap.json +9 -0
  7. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/moonep_async_overlap.txt +416 -0
  8. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/moonep_async_overlap_summary.txt +1 -0
  9. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/run.env +105 -0
  10. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/train.log +1139 -0
  11. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/verify_trace_ranks.log +6 -0
  12. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/results.tsv +1 -0
  13. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/trace_extension_build.log +58 -0
  14. granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/trace_results.tsv +2 -0
.gitattributes CHANGED
@@ -117,3 +117,4 @@ granularity_next_b200/tierb_models/traces/KIMI_K2_L2/ultraep/tierbtrace/tierbtra
117
  granularity_next_b200/tierb_models/traces/KIMI_K2_L2/moonep/tierbtrace/tierbtrace_gbs8/KIMI_K2_L2__moonep__matched__seq4096__gbs8__capnone__async/nsys/KIMI_K2_L2__moonep__matched__seq4096__gbs8__capnone__async.nsys-rep filter=lfs diff=lfs merge=lfs -text
118
  granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/echo/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__echo__matched__seq4096__gbs8__cap1p08984375/nsys/OLMOE_1B_7B_L16__echo__matched__seq4096__gbs8__cap1p08984375.nsys-rep filter=lfs diff=lfs merge=lfs -text
119
  granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/ultraep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__ultraep__matched__seq4096__gbs8__cap1p08984375/nsys/OLMOE_1B_7B_L16__ultraep__matched__seq4096__gbs8__cap1p08984375.nsys-rep filter=lfs diff=lfs merge=lfs -text
 
 
117
  granularity_next_b200/tierb_models/traces/KIMI_K2_L2/moonep/tierbtrace/tierbtrace_gbs8/KIMI_K2_L2__moonep__matched__seq4096__gbs8__capnone__async/nsys/KIMI_K2_L2__moonep__matched__seq4096__gbs8__capnone__async.nsys-rep filter=lfs diff=lfs merge=lfs -text
118
  granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/echo/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__echo__matched__seq4096__gbs8__cap1p08984375/nsys/OLMOE_1B_7B_L16__echo__matched__seq4096__gbs8__cap1p08984375.nsys-rep filter=lfs diff=lfs merge=lfs -text
119
  granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/ultraep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__ultraep__matched__seq4096__gbs8__cap1p08984375/nsys/OLMOE_1B_7B_L16__ultraep__matched__seq4096__gbs8__cap1p08984375.nsys-rep filter=lfs diff=lfs merge=lfs -text
120
+ granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.nsys-rep filter=lfs diff=lfs merge=lfs -text
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/manifest.txt ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ timestamp: 2026-09-21T09:05:26+00:00
2
+ host: sp-aiinfra-interactive-aghosh-8-mb99k-8xb200-moe-cg-0
3
+ gpu: NVIDIA B200 x8
4
+ driver: 580.95.05
5
+ study_scripts_branch: granularity-moonep
6
+ study_scripts_commit: d29f657
7
+ campaign: moonep
8
+ backend: moonep
9
+ phase: tierbtrace
10
+ gbs: 8
11
+ seq: 4096
12
+ configs: G1 G2 G3 G4
13
+ capacity: 1.25
14
+ repeats: 3
15
+ nsys: 1
16
+ megatron_repo: /workspace/Megatron-MoE-Echo
17
+ megatron_branch_required: moonep_backend_async_grad_reduce
18
+ megatron_branch_actual: moonep_backend_async_grad_reduce
19
+ megatron_commit: 2e68c8ad
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/pip_freeze.txt ADDED
@@ -0,0 +1,317 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ absl-py==2.3.0
2
+ aiohappyeyeballs @ file:///rapids/aiohappyeyeballs-2.6.1-py3-none-any.whl#sha256=f349ba8f4b75cb25c99c5c2d84e997e485204d2902a9597802b0371f09331fb8
3
+ aiohttp @ file:///rapids/aiohttp-3.12.7-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=94f98e0e5a49f89b252e115844f756c04fc8050f38252a32a3dd994ce8121f10
4
+ aiosignal @ file:///rapids/aiosignal-1.3.2-py2.py3-none-any.whl#sha256=45cde58e409a301715980c2b01d0c28bdde3770d8290b5eb2173759d9acb31a5
5
+ annotated-types==0.7.0
6
+ anyio==4.9.0
7
+ apex @ file:///opt/pytorch/apex/dist/apex-0.1-cp312-cp312-linux_x86_64.whl#sha256=a1508e4a386362e9c8f46c92ccada9110e421212b964c694235d0fd5a224b170
8
+ argon2-cffi==25.1.0
9
+ argon2-cffi-bindings==21.2.0
10
+ arrow==1.3.0
11
+ asciitree @ file:///rapids/asciitree-0.3.3-py3-none-any.whl#sha256=3e507ccfaeb417823158091fdb9d26b0205f6a59ba680b479a71db65c73d01d4
12
+ asttokens==3.0.0
13
+ astunparse==1.6.3
14
+ async-lru==2.0.5
15
+ attrs==25.3.0
16
+ audioread==3.0.1
17
+ babel==2.17.0
18
+ beautifulsoup4==4.13.4
19
+ black==25.1.0
20
+ bleach==6.2.0
21
+ blis==0.7.11
22
+ cachetools @ file:///rapids/cachetools-6.0.0-py3-none-any.whl#sha256=82e73ba88f7b30228b5507dce1a1f878498fc669d972aef2dde4f3a3c24f103e
23
+ catalogue==2.0.10
24
+ certifi==2025.4.26
25
+ cffi==1.17.1
26
+ charset-normalizer==3.4.2
27
+ click==8.5.0
28
+ cloudpathlib==0.21.1
29
+ cloudpickle @ file:///rapids/cloudpickle-3.1.1-py3-none-any.whl#sha256=c8c5a44295039331ee9dad40ba100a9c7297b6f988e50e87ccdf3765a668350e
30
+ cmake==3.31.6
31
+ comm==0.2.2
32
+ confection==0.1.5
33
+ contourpy==1.3.2
34
+ cuda-bindings @ file:///rapids/cuda_bindings-12.9.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=ff0e28d1e34758654b9c961e1f55e4786e49aee6a4dbceaf3cc24c46c672df7e
35
+ cuda-python @ file:///rapids/cuda_python-12.9.0-py3-none-any.whl#sha256=926acba49b2c0a0374c61b7c98f337c085199cf51cdfe4d6423c4129c20547a7
36
+ cudf @ file:///rapids/cudf-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=b779a8e8c0e8577b9b560895958496da211bfac49cdc4f11a49e8d18cb82e348
37
+ cudf-polars @ file:///rapids/cudf_polars-25.4.0-py3-none-any.whl#sha256=227b6c5147e93ea86282c8f7219118f916769fa2ca0cdd931e97a148aa66b869
38
+ cugraph @ file:///rapids/cugraph-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=b837c2dc2f7b235c702e4f8df64479b8fcbb342f848516d5faf0cdc4b5f31f29
39
+ cugraph-service-client @ file:///rapids/cugraph_service_client-25.4.0-py3-none-any.whl#sha256=7e568c3e41154b9752f5173555ffcc282ac52d72343feeb79653b2e2b1232470
40
+ cugraph-service-server @ file:///rapids/cugraph_service_server-25.4.0-py3-none-any.whl#sha256=a2830dcffafcccb16e12ebf85b98f8f247869d38c01ab4dd8664a1e7439d31be
41
+ cuml @ file:///rapids/cuml-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=1e90a464a18b094e24797f1676688c5d7696c4cf97c934504214a53d5aa09aa5
42
+ cupy-cuda12x @ file:///rapids/cupy_cuda12x-13.3.0-cp312-cp312-linux_x86_64.whl#sha256=6a2dfa0aa1e8189a10d1534276c08f1d2f73b747fd28bfbe7b9f4ef66d9af73f
43
+ cuvs @ file:///rapids/cuvs-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=b12f6b9f80e053a3b38542c2947d2516e468c02cad54a13d8b1a4a61d5aa7914
44
+ cycler==0.12.1
45
+ cymem==2.0.11
46
+ Cython==3.1.1
47
+ dask @ file:///rapids/dask-2025.2.0-py3-none-any.whl#sha256=f0fdeef6ceb0a06569d456c9e704f220f7f54e80f3a6ea42ab98cea6bc642b6e
48
+ dask-cuda @ file:///rapids/dask_cuda-25.4.0-py3-none-any.whl#sha256=442d87da0a95d00ca2a262b64c76e785a32ad905c82000fa2227932f81390e19
49
+ dask-cudf @ file:///rapids/dask_cudf-25.4.0-py3-none-any.whl#sha256=5e7cea0b26daea17da6e0bb457435e24460f76ba740358a3a18e239d6015c95c
50
+ datasets==5.0.1
51
+ debugpy==1.8.14
52
+ decorator==5.2.1
53
+ -e git+https://github.com/deepseek-ai/DeepEP.git@34152ae28f80bcc3ee38d7a12cb2ad87cfd4ea72#egg=deep_ep
54
+ defusedxml==0.7.1
55
+ dill==0.4.1
56
+ distributed @ file:///rapids/distributed-2025.2.0-py3-none-any.whl#sha256=368462084be47092d0a835f6c64c9b852da64dc88302b76d5f93bd5131228c67
57
+ distributed-ucxx @ file:///rapids/distributed_ucxx-0.43.0-py3-none-any.whl#sha256=94f60d0ccb0c94ba8dbfa1eaf455874d18aaab17e2ae218c38590e9fdc2d721d
58
+ dm-tree==0.1.9
59
+ einops==0.8.1
60
+ execnet==2.1.1
61
+ executing==2.2.0
62
+ expecttest==0.3.0
63
+ fasteners @ file:///rapids/fasteners-0.19-py3-none-any.whl#sha256=758819cb5d94cdedf4e836988b74de396ceacb8e2794d21f82d131fd9ee77237
64
+ fastjsonschema==2.21.1
65
+ fastrlock @ file:///rapids/fastrlock-0.8.3-cp312-cp312-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_28_x86_64.whl#sha256=5f13ec08f1adb1aa916c384b05ecb7dbebb8df9ea81abd045f60941c6283a670
66
+ filelock==3.18.0
67
+ flash_attn @ file:///opt/pytorch/flash_attn-2.7.4.post1-cp312-cp312-linux_x86_64.whl#sha256=a10455d6c56b387785b7413fd0bbcb392b42b3d4f4b88fda5d86cc8a9ce9182a
68
+ fonttools==4.58.1
69
+ fqdn==1.5.1
70
+ frozenlist @ file:///rapids/frozenlist-1.6.0-cp312-cp312-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=4e1be9111cb6756868ac242b3c2bd1f09d9aea09846e4f5c23715e7afb647103
71
+ fsspec==2025.5.1
72
+ gast==0.6.0
73
+ grpcio==1.62.1
74
+ h11==0.16.0
75
+ hf-xet==1.6.0
76
+ httpcore==1.0.9
77
+ httpx==0.28.1
78
+ huggingface_hub==1.32.0
79
+ hypothesis==6.130.8
80
+ idna==3.10
81
+ importlib_metadata==8.7.0
82
+ iniconfig==2.1.0
83
+ intel-openmp==2021.4.0
84
+ ipykernel==6.29.5
85
+ ipython==9.3.0
86
+ ipython_pygments_lexers==1.1.1
87
+ isoduration==20.11.0
88
+ isort==6.0.1
89
+ jedi==0.19.2
90
+ Jinja2==3.1.6
91
+ joblib==1.5.1
92
+ json5==0.12.0
93
+ jsonpointer==3.0.0
94
+ jsonschema==4.24.0
95
+ jsonschema-specifications==2025.4.1
96
+ jupyter-events==0.12.0
97
+ jupyter-lsp==2.2.5
98
+ jupyter_client==8.6.3
99
+ jupyter_core==5.8.1
100
+ jupyter_server==2.16.0
101
+ jupyter_server_terminals==0.5.3
102
+ jupyterlab==4.4.3
103
+ jupyterlab_code_formatter==3.0.2
104
+ jupyterlab_pygments==0.3.0
105
+ jupyterlab_server==2.27.3
106
+ jupyterlab_tensorboard_pro @ file:///builder/jupyterlab_tensorboard_pro-4.0.0-py2.py3-none-any.whl#sha256=6bfcb5ced3358eb72150f4b2ab91312faef2f0ca14d845e8379d8fa52df6902e
107
+ jupytext==1.17.2
108
+ kiwisolver==1.4.8
109
+ kvikio @ file:///rapids/kvikio-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=8b40fd8ea9e4a2974c04b50b7fc26b06094ac807f9a233281d0bafbaa71b1d85
110
+ langcodes==3.5.0
111
+ language_data==1.3.0
112
+ lazy_loader==0.4
113
+ libcudf @ file:///rapids/libcudf-25.4.0-py3-none-linux_x86_64.whl#sha256=238320b4958c1737bba2a609673e2539a78216b811b5498a68e00395d9b021a4
114
+ libcugraph @ file:///rapids/libcugraph-25.4.0-py3-none-linux_x86_64.whl#sha256=29085e901b2f7569ff2653bb1355fe6fad20f2fcaa585f5649da9867d8df766d
115
+ libcuml @ file:///rapids/libcuml-25.4.0-py3-none-linux_x86_64.whl#sha256=8e2e444c72b0e995e15a48da37ff12fb410a47a2baa69ba86ccc9b1a3db91e8c
116
+ libcuvs @ file:///rapids/libcuvs-25.4.0-py3-none-linux_x86_64.whl#sha256=e2f8db61c92b769c590dd752f43439ddd62b2f6592a3c337362b6c1ef220c3b9
117
+ libkvikio @ file:///rapids/libkvikio-25.4.0-py3-none-linux_x86_64.whl#sha256=da8a5713b05fc478bf40529a3e0fb61a2d605bb5a70ca4d461f96a6005dbe8b7
118
+ libraft @ file:///rapids/libraft-25.4.0-py3-none-linux_x86_64.whl#sha256=b921dec005aa4a321dd156776a18376feb47c96a5b67eec125c8a8e19edd1e24
119
+ librmm @ file:///rapids/librmm-25.4.0-py3-none-linux_x86_64.whl#sha256=ff951811bda729d4fa1109ab906225ae1f489a6ec2eeab5d1b674cc34bd4422f
120
+ librmm-cu12 @ file:///rapids/librmm_cu12-25.4.0-py3-none-any.whl#sha256=29296ef2b0e90679a0db280e86015be12d6ab9f4e62a181aa085f62dbadd06b3
121
+ librosa==0.11.0
122
+ libucxx @ file:///rapids/libucxx-0.43.0-py3-none-linux_x86_64.whl#sha256=f15849d7ac2fb68f0fd8998838a13321ddcfb0c594f7e6cbc401a3bcc4750566
123
+ lightning-thunder @ file:///opt/pytorch/lightning-thunder
124
+ lightning-utilities==0.14.3
125
+ lintrunner==0.12.7
126
+ llvmlite==0.42.0
127
+ locket @ file:///rapids/locket-1.0.0-py2.py3-none-any.whl#sha256=b6c819a722f7b6bd955b80781788e4a66a55628b858d347536b7e81325a3a5e3
128
+ looseversion==1.3.0
129
+ marisa-trie==1.2.1
130
+ Markdown==3.8
131
+ markdown-it-py==3.0.0
132
+ MarkupSafe==3.0.2
133
+ matplotlib==3.10.3
134
+ matplotlib-inline==0.1.7
135
+ mdit-py-plugins==0.4.2
136
+ mdurl==0.1.2
137
+ -e git+ssh://git@github.com/sriram-17-17/my-Megatron-MoE-Echo.git@2e68c8ad6bfb5bac5d6eb962fa7d4fbf0a5fb095#egg=megatron_core
138
+ mistune==3.1.3
139
+ mkl==2021.1.1
140
+ mkl-devel==2021.1.1
141
+ mkl-include==2021.1.1
142
+ ml_dtypes==0.6.0
143
+ mock==5.2.0
144
+ -e git+ssh://git@github.com/sriram-17-17/MoonEP.git@d8bc6667fad7177e6e09b637d773fa67f08bae93#egg=moonep
145
+ mpmath==1.3.0
146
+ msgpack==1.1.0
147
+ multidict @ file:///rapids/multidict-6.4.4-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=a920f9cf2abdf6e493c519492d892c362007f113c94da4c239ae88429835bad1
148
+ multiprocess==0.70.19
149
+ murmurhash==1.0.13
150
+ mypy_extensions==1.1.0
151
+ nbclient==0.10.2
152
+ nbconvert==7.16.6
153
+ nbformat==5.10.4
154
+ nest-asyncio==1.6.0
155
+ networkx==3.5
156
+ ninja==1.11.1.4
157
+ notebook==7.4.3
158
+ notebook_shim==0.2.4
159
+ numba==0.59.1
160
+ numba-cuda @ file:///rapids/numba_cuda-0.4.0-py3-none-any.whl#sha256=5b95a5f42068771e37142944d44fe844fa87f1b2f89051461a4cb0af91bcc28a
161
+ numcodecs @ file:///rapids/numcodecs-0.13.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=da2230484e6102e5fa3cc1a5dd37ca1f92dfbd183d91662074d6f7574e3e8f53
162
+ numpy==1.26.4
163
+ nvdlfw_inspect @ git+https://github.com/NVIDIA/nvidia-dlfw-inspect.git@6b60eb0e675606fb2fbbfcb12667ecbec75eaf87
164
+ nvfuser==0.2.27a0+9bf5aca
165
+ nvidia-cuda-cccl-cu12==12.9.27
166
+ nvidia-cudnn-frontend @ file:///opt/pytorch/lightning-thunder/tmp/cudnn_frontend
167
+ nvidia-cutlass-dsl==4.4.2
168
+ nvidia-cutlass-dsl-libs-base==4.4.2
169
+ nvidia-dali-cuda120==1.50.0
170
+ nvidia-ml-py @ file:///rapids/nvidia_ml_py-12.575.51-py3-none-any.whl#sha256=eb8641800d98ce40a22f479873f34b482e214a7e80349c63be51c3919845446e
171
+ nvidia-modelopt==0.29.0
172
+ nvidia-modelopt-core==0.29.0
173
+ nvidia-nvcomp-cu12==4.2.0.14
174
+ nvidia-nvimgcodec-cu12==0.5.0.13
175
+ nvidia-nvjpeg-cu12==12.4.0.16
176
+ nvidia-nvjpeg2k-cu12==0.8.1.40
177
+ nvidia-nvshmem-cu12==3.7.2
178
+ nvidia-nvtiff-cu12==0.5.0.67
179
+ nvidia-resiliency-ext==0.4.0
180
+ nvtx @ file:///rapids/nvtx-0.2.11-cp312-cp312-linux_x86_64.whl#sha256=73521e37cba42b4db0b3cd68fc5eb814e55c500a61eeed6edcdf6746315ac53e
181
+ nx-cugraph @ file:///rapids/nx_cugraph-25.4.0-py3-none-any.whl#sha256=e625444486cc899a77bd48e7da76a2edd95c0006dd95fbf323745feae7f33111
182
+ onnx @ file:///opt/pytorch/pytorch/third_party/onnx
183
+ onnx-ir==1.0.0
184
+ onnxscript==0.7.2
185
+ opt_einsum==3.4.0
186
+ optree==0.16.0
187
+ overrides==7.7.0
188
+ packaging==23.2
189
+ pandas @ file:///rapids/pandas-2.2.3-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=fffb8ae78d8af97f849404f21411c95062db1496aeb3e56f146f0355c9989319
190
+ pandocfilters==1.5.1
191
+ parso==0.8.4
192
+ partd @ file:///rapids/partd-1.4.2-py3-none-any.whl#sha256=978e4ac767ec4ba5b86c6eaa52e5a2a3bc748a2ca839e8cc798f1cc6ce6efb0f
193
+ pathspec==0.12.1
194
+ pexpect==4.9.0
195
+ pillow==11.2.1
196
+ platformdirs==4.3.8
197
+ pluggy==1.6.0
198
+ ply @ file:///rapids/ply-3.11-py2.py3-none-any.whl#sha256=096f9b8350b65ebd2fd1346b12452efe5b9607f7482813ffca50c22722a807ce
199
+ polars @ file:///rapids/polars-1.25.2-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=f7fcbb4f476784384ccda48757fca4e8c2e2c5a0a3aef3717aaf56aee4e30e09
200
+ polygraphy==0.49.20
201
+ pooch==1.8.2
202
+ preshed==3.0.10
203
+ prometheus_client==0.22.1
204
+ prompt_toolkit==3.0.51
205
+ propcache @ file:///rapids/propcache-0.3.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=563f9d8c03ad645597b8d010ef4e9eab359faeb11a0a2ac9f7b4bc8c28ebef25
206
+ protobuf==4.24.4
207
+ psutil==7.0.0
208
+ ptyprocess==0.7.0
209
+ PuLP==3.2.1
210
+ pure_eval==0.2.3
211
+ pyarrow==25.0.1
212
+ pybind11==2.13.6
213
+ pybind11_global==2.13.6
214
+ pycocotools @ git+https://github.com/nvidia/cocoapi.git@08350a157b41f0755a536441b5abec89ab934c5b#subdirectory=PythonAPI
215
+ pycparser==2.22
216
+ pydantic==2.11.5
217
+ pydantic_core==2.33.2
218
+ Pygments==2.19.1
219
+ pylibcudf @ file:///rapids/pylibcudf-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=401b8bfd022d06f84e120779ca7c3d01a28b68762aaa8f7873b10d6b1975cf6a
220
+ pylibcugraph @ file:///rapids/pylibcugraph-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=061c7389a88c815b54fd5d3a3584fb87bf693e5602f5be335260d0ba575ba557
221
+ pylibcugraphops @ file:///rapids/pylibcugraphops-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=a25ef4b22166c80718d5c7e5bddb755e81a2fe340fa5becd0ac36ec3edfb208a
222
+ pylibraft @ file:///rapids/pylibraft-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=f2ecd23fe51f9c2caa87bbbe85cb32fcaff4ecfb52ceb9a9991bc50f82782c2e
223
+ pylibwholegraph @ file:///rapids/pylibwholegraph-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=b82913d1e38bda76d8c272fbd8cc9072d987209d3d47cc93a464a37c2925aca2
224
+ pynvjitlink @ file:///rapids/pynvjitlink-0.3.0-cp312-cp312-linux_x86_64.whl#sha256=066caf1314730643d3b48c9847bd65789defdd6a7c78993176eb3d8bbe3f4919
225
+ pynvml @ file:///rapids/pynvml-12.0.0-py3-none-any.whl#sha256=fdff84b62a27dbe98e08e1a647eb77342bef1aebe0878bcd15e99a83fcbecb9e
226
+ pyparsing==3.2.3
227
+ pytest==8.1.1
228
+ pytest-flakefinder==1.1.0
229
+ pytest-rerunfailures==15.1
230
+ pytest-shard==0.1.2
231
+ pytest-xdist==3.7.0
232
+ python-dateutil==2.9.0.post0
233
+ python-hostlist==2.2.1
234
+ python-json-logger==3.3.0
235
+ pytorch-triton @ file:///tmp/dist/pytorch_triton-3.3.0%2Bgit96316ce52.nvinternal-cp312-cp312-linux_x86_64.whl#sha256=6d6921772b9dfd0243c055e0d054de42a113e92cabcfdd1d3e80a5ad9c08cf7d
236
+ pytz @ file:///rapids/pytz-2023.4-py2.py3-none-any.whl#sha256=f90ef520d95e7c46951105338d918664ebfd6f1d995bd7d153127ce90efafa6a
237
+ PyYAML==6.0.2
238
+ pyzmq==26.4.0
239
+ raft-dask @ file:///rapids/raft_dask-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=36cf37ba6ce35454c4729c2cb597a360f22780eefb11d514a16a07137da5eb6c
240
+ rapids-dask-dependency @ file:///rapids/rapids_dask_dependency-25.4.0a0-py3-none-any.whl#sha256=484faca626e7760112a7d8ad3426f5bdd720b609006d8dca87c2607076a348bd
241
+ rapids-logger @ file:///rapids/rapids_logger-0.1.18-py3-none-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=3617b2d86291ac6cb957542895f20fc9737ace29773ca90337e6fb04c73bd120
242
+ referencing==0.36.2
243
+ regex==2024.11.6
244
+ requests==2.32.3
245
+ rfc3339-validator==0.1.4
246
+ rfc3986-validator==0.1.1
247
+ rich==14.0.0
248
+ rmm @ file:///rapids/rmm-25.4.0-cp312-cp312-linux_x86_64.whl#sha256=1824bcab3836296666980a997ddcb9c4bca24dfb873012cfa0a7ecea583ca712
249
+ rpds-py==0.25.1
250
+ safetensors==0.5.3
251
+ scikit-learn==1.6.1
252
+ scipy==1.15.3
253
+ Send2Trash==1.8.3
254
+ setuptools==78.1.1
255
+ shellingham==1.5.4
256
+ six==1.16.0
257
+ smart-open==7.1.0
258
+ sniffio==1.3.1
259
+ sortedcontainers==2.4.0
260
+ soundfile==0.13.1
261
+ soupsieve==2.7
262
+ soxr==0.5.0.post1
263
+ spacy==3.7.5
264
+ spacy-legacy==3.0.12
265
+ spacy-loggers==1.0.5
266
+ srsly==2.5.1
267
+ stack-data==0.6.3
268
+ sympy==1.14.0
269
+ tabulate==0.9.0
270
+ tbb==2021.13.1
271
+ tblib @ file:///rapids/tblib-3.1.0-py3-none-any.whl#sha256=670bb4582578134b3d81a84afa1b016128b429f3d48e6cbbaecc9d15675e984e
272
+ tensorboard==2.16.2
273
+ tensorboard-data-server==0.7.2
274
+ tensorrt @ file:///workspace/TensorRT-10.11.0.33/python/tensorrt-10.11.0.33-cp312-none-linux_x86_64.whl#sha256=c2f18f72758a3d54a71901495c8fda9e7094a8bcdd2a5b6b58967fbf05733b3a
275
+ terminado==0.18.1
276
+ thinc==8.2.5
277
+ threadpoolctl==3.6.0
278
+ thriftpy2 @ file:///rapids/thriftpy2-0.5.2-cp312-cp312-linux_x86_64.whl#sha256=c0c7d939bdc5ccf7739154959938445d5fdf72f1d6e12bd563947a36f67e2147
279
+ tinycss2==1.4.0
280
+ toolz @ file:///rapids/toolz-1.0.0-py3-none-any.whl#sha256=292c8f1c4e7516bf9086f8850935c799a874039c8bcf959d47b600e4c44a6236
281
+ torch @ file:///opt/transfer/torch-2.8.0a0%2B5228986c39.nv25.6-29946474-cp312-cp312-linux_x86_64.whl#sha256=7a3c9ff5288d8102a31296b3d62491f3f47c8820a5c9aaefc7dc0fe20e1e8ad5
282
+ torch_tensorrt @ file:///opt/pytorch/torch_tensorrt/dist/torch_tensorrt-2.8.0a0-cp312-cp312-linux_x86_64.whl#sha256=26f4d7a04e7e6e017a5b917b8b98b83720bb2bb0195f48d25f52d567d286bfe1
283
+ torchao @ file:///opt/pytorch/ao
284
+ torchprofile==0.0.4
285
+ torchvision @ file:///opt/transfer/torchvision-0.22.0a0%2B95f10a4e-29946474-cp312-cp312-manylinux_2_28_x86_64.whl#sha256=746a3c674cc61680a86f4a14b374b21af1b5c21f629c42765b40c9494f6b234d
286
+ tornado==6.5.1
287
+ tqdm==4.67.1
288
+ traitlets==5.14.3
289
+ transformer_engine @ file:///pvcdata/aiinfra/b-aghosh/wheels/transformer_engine-2.10.0.dev0%2B10f6128d-cp312-cp312-linux_x86_64.whl#sha256=35cf5f4b83b2baad864e12e512bb422e0929409ac05610812fa521d4942d0189
290
+ treelite @ file:///rapids/treelite-4.4.1-py3-none-linux_x86_64.whl#sha256=174741724205540a1a5b37da81483aeff55d89548f42b3461cf07acac7ae96eb
291
+ typer==0.16.0
292
+ types-dataclasses==0.6.6
293
+ types-python-dateutil==2.9.0.20250516
294
+ typing-inspection==0.4.1
295
+ typing_extensions==4.14.0
296
+ tzdata @ file:///rapids/tzdata-2025.2-py2.py3-none-any.whl#sha256=1a403fada01ff9221ca8044d701868fa132215d84beb92242d9acd2147f667a8
297
+ ucx-py @ file:///rapids/ucx_py-0.43.0-cp312-cp312-linux_x86_64.whl#sha256=4a452fd8d8c983283ada3d49086e3657dc2ee94fbad08efdd708d15ab6e0b73a
298
+ ucxx @ file:///rapids/ucxx-0.43.0-cp312-cp312-linux_x86_64.whl#sha256=f1b4e55cb45c66fe88a2d8f9c2b87751165f3cfec21fbdc9c47c4ddff75e8a5f
299
+ -e git+https://github.com/Dots-Infra/UltraEP.git@94cab099b44fffa99a82fea99e7c12d89cf65e4f#egg=ultra_ep
300
+ uri-template==1.3.0
301
+ urllib3==2.0.7
302
+ wasabi==1.1.3
303
+ wcwidth==0.2.13
304
+ weasel==0.4.1
305
+ webcolors==24.11.1
306
+ webencodings==0.5.1
307
+ websocket-client==1.8.0
308
+ Werkzeug==3.1.3
309
+ wheel==0.45.1
310
+ wrapt==1.17.2
311
+ xdoctest==1.0.2
312
+ xgboost @ file:///rapids/xgboost-2.1.4-py3-none-manylinux_2_39_x86_64.whl#sha256=25eeca27f4048fc491940b3a0b4b5782ee48c33cb3aec059cef71f8da542930b
313
+ xxhash==4.0.1
314
+ yarl @ file:///rapids/yarl-1.20.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl#sha256=bf099e2432131093cc611623e0b0bcc399b8cddd9a91eded8bfb50402ec35018
315
+ zarr @ file:///rapids/zarr-2.18.7-py3-none-any.whl#sha256=ac3dc4033e9ae4e9d7b5e27c97ea3eaf1003cc0a07f010bd83d5134bf8c4b223
316
+ zict @ file:///rapids/zict-3.0.0-py2.py3-none-any.whl#sha256=5796e36bd0e0cc8cf0fbc1ace6a68912611c1dbd74750a3f3026b9b9d6a327ae
317
+ zipp==3.22.0
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/provenance.txt ADDED
@@ -0,0 +1,90 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ==============================================================
2
+ PROVENANCE: moonep__tierbtrace__gbs8
3
+ ==============================================================
4
+ captured_at : 2026-09-21T09:05:12+00:00
5
+ hostname : sp-aiinfra-interactive-aghosh-8-mb99k-8xb200-moe-cg-0
6
+ kernel : 6.8.0-1043-nvidia
7
+ cwd : /workspace/blackwell_run_scripts
8
+ command : run_ladder.sh --phase tierbtrace --gbs 8 --out /workspace/granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace --nsys
9
+
10
+ ---- GPU / driver --------------------------------------------
11
+ 0, NVIDIA B200, 10.0, 183359 MiB, 580.95.05
12
+ 1, NVIDIA B200, 10.0, 183359 MiB, 580.95.05
13
+ 2, NVIDIA B200, 10.0, 183359 MiB, 580.95.05
14
+ 3, NVIDIA B200, 10.0, 183359 MiB, 580.95.05
15
+ 4, NVIDIA B200, 10.0, 183359 MiB, 580.95.05
16
+ 5, NVIDIA B200, 10.0, 183359 MiB, 580.95.05
17
+ 6, NVIDIA B200, 10.0, 183359 MiB, 580.95.05
18
+ 7, NVIDIA B200, 10.0, 183359 MiB, 580.95.05
19
+ persistence/clocks:
20
+ 0, 120 MHz, 1965 MHz, 1000.00 W, 29
21
+ 1, 1965 MHz, 1965 MHz, 1000.00 W, 35
22
+ 2, 120 MHz, 1965 MHz, 1000.00 W, 34
23
+ 3, 120 MHz, 1965 MHz, 1000.00 W, 30
24
+ 4, 1965 MHz, 1965 MHz, 1000.00 W, 30
25
+ 5, 120 MHz, 1965 MHz, 1000.00 W, 33
26
+ 6, 120 MHz, 1965 MHz, 1000.00 W, 33
27
+ 7, 120 MHz, 1965 MHz, 1000.00 W, 29
28
+
29
+ NVLink topology:
30
+ GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 NIC7 NIC8 CPU Affinity NUMA Affinity GPU NUMA ID
31
+ GPU0 X NV18 NV18 NV18 NV18 NV18 NV18 NV18 PIX NODE SYS SYS SYS SYS SYS SYS NODE 0-27 0 N/A
32
+ GPU1 NV18 X NV18 NV18 NV18 NV18 NV18 NV18 NODE PIX SYS SYS SYS SYS SYS SYS NODE 0-27 0 N/A
33
+ GPU2 NV18 NV18 X NV18 NV18 NV18 NV18 NV18 SYS SYS PIX NODE SYS SYS SYS SYS SYS 28-55 1 N/A
34
+ GPU3 NV18 NV18 NV18 X NV18 NV18 NV18 NV18 SYS SYS NODE PIX SYS SYS SYS SYS SYS 28-55 1 N/A
35
+ GPU4 NV18 NV18 NV18 NV18 X NV18 NV18 NV18 SYS SYS SYS SYS PIX NODE SYS SYS SYS 56-83 2 N/A
36
+ GPU5 NV18 NV18 NV18 NV18 NV18 X NV18 NV18 SYS SYS SYS SYS NODE PIX SYS SYS SYS 56-83 2 N/A
37
+ GPU6 NV18 NV18 NV18 NV18 NV18 NV18 X NV18 SYS SYS SYS SYS SYS SYS PIX NODE SYS 84-111 3 N/A
38
+ GPU7 NV18 NV18 NV18 NV18 NV18 NV18 NV18 X SYS SYS SYS SYS SYS SYS NODE PIX SYS 84-111 3 N/A
39
+ NIC0 PIX NODE SYS SYS SYS SYS SYS SYS X NODE SYS SYS SYS SYS SYS SYS NODE
40
+ NIC1 NODE PIX SYS SYS SYS SYS SYS SYS NODE X SYS SYS SYS SYS SYS SYS NODE
41
+ NIC2 SYS SYS PIX NODE SYS SYS SYS SYS SYS SYS X NODE SYS SYS SYS SYS SYS
42
+ NIC3 SYS SYS NODE PIX SYS SYS SYS SYS SYS SYS NODE X SYS SYS SYS SYS SYS
43
+ NIC4 SYS SYS SYS SYS PIX NODE SYS SYS SYS SYS SYS SYS X NODE SYS SYS SYS
44
+ unavailable
45
+
46
+ ---- toolchain -----------------------------------------------
47
+ nvcc : release 12.9, V12.9.86
48
+ gcc : gcc (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0
49
+ python : 3.12.3
50
+ torch : 2.8.0a0+5228986c39.nv25.06 cuda=12.9 nccl=2.27.3
51
+ torch arch : ['sm_75', 'sm_80', 'sm_86', 'sm_90', 'sm_100', 'sm_120', 'compute_120']
52
+ transformer_engine: 2.10.0.dev0+10f6128d [/usr/local/lib/python3.12/dist-packages/transformer_engine/__init__.py]
53
+ deep_ep : 1.2.1+34152ae [/workspace/DeepEP/deep_ep/__init__.py]
54
+ ultra_ep : 1.0.0 [/workspace/UltraEP/ultra_ep/__init__.py]
55
+ moonep : 0.0.1 [/workspace/MoonEP/moonep/__init__.py]
56
+ megatron : 0.14.0rc5 [None]
57
+ flash_attn : 2.7.4.post1 [/usr/local/lib/python3.12/dist-packages/flash_attn/__init__.py]
58
+ numpy : 1.26.4 [/usr/local/lib/python3.12/dist-packages/numpy/__init__.py]
59
+ huggingface_hub: 1.32.0 [/usr/local/lib/python3.12/dist-packages/huggingface_hub/__init__.py]
60
+
61
+ ---- repositories --------------------------------------------
62
+ Megatron-MoE-Echo moonep_backend_async_grad_reduce 2e68c8ad6bfb
63
+ Megatron-UltraEP ultraep-bracing-regions-through-kernels 03d57ef7d476
64
+ UltraEP main 94cab099b44f
65
+ DeepEP HEAD 34152ae28f80 (+3 untracked: build output)
66
+ MoonEP megatron-ddp-symmetric-params d8bc6667fad7
67
+ blackwell_run_scripts granularity-moonep d29f6570cdce
68
+
69
+ ---- environment ---------------------------------------------
70
+ CUDA_ARCH_LIST=7.5 8.0 8.6 9.0 10.0 12.0
71
+ CUDA_DRIVER_VERSION=575.57.08
72
+ CUDA_HOME=/usr/local/cuda
73
+ CUDA_MODULE_LOADING=LAZY
74
+ CUDA_VERSION=12.9.1.010
75
+ HF_REMOTE_ROOT=granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16
76
+ NCCL_VERSION=2.27.3
77
+ NVSHMEM_VERSION=3.2.5
78
+ PYTORCH_BUILD_NUMBER=0
79
+ PYTORCH_BUILD_VERSION=2.8.0a0+5228986
80
+ PYTORCH_HOME=/opt/pytorch/pytorch
81
+ PYTORCH_VERSION=2.8.0a0+5228986
82
+ TORCH_ALLOW_TF32_CUBLAS_OVERRIDE=1
83
+ TORCH_CUDA_ARCH_LIST=7.5 8.0 8.6 9.0 10.0 12.0+PTX
84
+ TORCH_NCCL_USE_COMM_NONBLOCKING=0
85
+
86
+ ---- shared library resolution -------------------------------
87
+ libnvshmem_host.so: not in ldconfig
88
+ libnccl.so: /lib/x86_64-linux-gnu/libnccl.so.2
89
+ lrwxrwxrwx 1 root root 20 Sep 21 04:57 /usr/local/lib/python3.12/dist-packages/nvidia/nvshmem/lib/libnvshmem_host.so -> libnvshmem_host.so.3
90
+ -rw-r--r-- 1 root root 179045744 Sep 21 04:54 /usr/local/lib/python3.12/dist-packages/nvidia/nvshmem/lib/libnvshmem_host.so.3
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.nsys-rep ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04e41ce6a8ee7878ae5a344f4dad734952f48af9bdaf25fba2be183d2bba8731
3
+ size 7288090
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/moonep_async_overlap.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "devices": 8,
3
+ "reduce_busy_ms_per_iteration": 2.612875,
4
+ "overlapped_ms_per_iteration": 2.330125,
5
+ "exposed_ms_per_iteration": 0.282625,
6
+ "overlap_pct": 89.17858680572166,
7
+ "minimum_overlap_pct": 0.1,
8
+ "status": "PASS"
9
+ }
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/moonep_async_overlap.txt ADDED
@@ -0,0 +1,416 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ === kernels on device 0, by total time ===
3
+ 26.953 ms elementwise_kernel
4
+ 20.968 ms device_kernel
5
+ 17.521 ms kernel_cutlass_kernel_moonepdispatchDispatchKernel_object_at__tensorptrbf16gmemalign16o40962048i64i64_tensorptrbf16gmemalign16o2785282048i64i64_tensorptri32gmemalign16o327681_tensorptri32_0
6
+ 15.802 ms kernel_cutlass_kernel_moonepcombineCombineKernel_object_at__tensorptrbf16gmemalign16o2785282048i64i64_tensorptrbf16gmemalign16o40962048i64i64_tensorptri32gmemalign16o327681_tensorptri32gm_0
7
+ 14.136 ms multi_tensor_apply_kernel
8
+ 13.031 ms vectorized_elementwise_kernel
9
+ 13.000 ms cast_mxfp8_2D_kernel
10
+ 10.454 ms ncclDevKernel_AllReduce_Sum_f32_RING_LL
11
+ 8.005 ms ncclDevKernel_ReduceScatter_Sum_f32_RING_LL
12
+ 6.272 ms unrolled_elementwise_kernel
13
+ 6.154 ms cudnn_generated_fort_native_sdpa_sm100_flash_bprop_f16_knob_31_128x128x128_1x4x1_kernel0_0
14
+ 4.095 ms ncclDevKernel_AllGather_RING_LL
15
+ 3.963 ms kernel_cutlass_kernel_moonepinter_rank_syncInterRankSyncKernel_object_at__tensorptri32gmemalign16o41943041_tensorptri32gmemalign16o10___0
16
+ 3.704 ms triton_poi_fused__to_copy_cat_0
17
+ REDUCE 3.474 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310722048i64i64_tensorptrf32gmemalign16o1310722048i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
18
+ 2.847 ms cudnn_generated_fort_native_sdpa_sm100_flash_fprop_f16_knob_7_128x128x128_4x1x1_kernel0_0
19
+ 1.971 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_TNT
20
+ 1.949 ms nvjet_qqsss_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NTT
21
+ 1.910 ms kernel_cutlass_kernel_moonepplanningPlanningKernel_object_at__tensorptri32_gmem_align16_o_641_tensorptri32_gmem_align16_o_327681_tensorptri32_gmem_align16_o_41943041_tensorptri32_gmem_ali_0
22
+ REDUCE 1.894 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310721024i64i64_tensorptrf32gmemalign16o1310721024i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
23
+ 1.858 ms kernel_cutlass_kernel_moonepcombine_prologueCombinePrologueKernel_object_at__tensorptrbf16gmemalign16o348002048i64i64_tensorptri32gmemalign16o1044001_tensorptri32gmemalign16o348001_tensor_0
24
+ 1.790 ms reduce_kernel
25
+ 1.758 ms CatArrayBatchedCopy_alignedK_contig
26
+ 1.725 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NNT
27
+ 1.714 ms _scatter_gather_elementwise_kernel
28
+ 1.636 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072204811i640_CopyAtom_ThrID10_TVLayoutSrc_0
29
+ 1.472 ms gatherTopK
30
+ 1.323 ms triton_red_fused__to_copy_mul_silu_sum_1
31
+ 1.251 ms fused_rope_forward_kernel
32
+ 1.242 ms multi_tensor_swizzle_row_scaling_kernel
33
+ 1.159 ms fused_rope_backward_kernel
34
+ 1.153 ms setGroupedGemmWgradArguments
35
+ 1.148 ms multi_tensor_swizzle_col_scaling_kernel
36
+ 1.089 ms triton_poi_fused__to_copy_mul_silu_0
37
+ 1.024 ms swizzle_col_scaling_kernel
38
+ 0.984 ms nvjet_tss_128x256_64x6_2x1_2cta_v_badd_NTT
39
+ 0.974 ms nvjet_tst_256x256_64x4_2x2_2cta_h_bz_NNT
40
+ 0.943 ms rmsnorm_bwd_tuned_kernel
41
+ 0.888 ms nvjet_tst_128x256_64x6_2x1_2cta_v_bz_TNT
42
+ 0.837 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072102411i640_CopyAtom_ThrID10_TVLayoutSrc_0
43
+
44
+ === OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.sqlite (device 0, 2 iteration(s)) ===
45
+ reduce_grad kernels : 64 (32/iter)
46
+ reduce_grad sum : 2.684 ms/iter
47
+ reduce_grad busy (union) : 2.684 ms/iter
48
+ ... overlapped with compute: 2.402 ms/iter ( 89.5 %)
49
+ ... exposed on critical path: 0.281 ms/iter ( 10.5 %)
50
+ other GPU work (union) : 97.827 ms/iter
51
+ traced wall : 126.187 ms/iter
52
+ --> time hidden by overlap : 2.402 ms/iter
53
+
54
+ === kernels on device 1, by total time ===
55
+ 26.922 ms elementwise_kernel
56
+ 20.799 ms device_kernel
57
+ 17.483 ms kernel_cutlass_kernel_moonepdispatchDispatchKernel_object_at__tensorptrbf16gmemalign16o40962048i64i64_tensorptrbf16gmemalign16o2785282048i64i64_tensorptri32gmemalign16o327681_tensorptri32_0
58
+ 16.354 ms kernel_cutlass_kernel_moonepcombineCombineKernel_object_at__tensorptrbf16gmemalign16o2785282048i64i64_tensorptrbf16gmemalign16o40962048i64i64_tensorptri32gmemalign16o327681_tensorptri32gm_0
59
+ 14.127 ms multi_tensor_apply_kernel
60
+ 13.043 ms cast_mxfp8_2D_kernel
61
+ 12.990 ms vectorized_elementwise_kernel
62
+ 12.477 ms ncclDevKernel_AllReduce_Sum_f32_RING_LL
63
+ 8.145 ms ncclDevKernel_ReduceScatter_Sum_f32_RING_LL
64
+ 6.293 ms unrolled_elementwise_kernel
65
+ 6.112 ms cudnn_generated_fort_native_sdpa_sm100_flash_bprop_f16_knob_31_128x128x128_1x4x1_kernel0_0
66
+ 4.486 ms ncclDevKernel_AllGather_RING_LL
67
+ 4.015 ms kernel_cutlass_kernel_moonepinter_rank_syncInterRankSyncKernel_object_at__tensorptri32gmemalign16o41943041_tensorptri32gmemalign16o10___0
68
+ 3.697 ms triton_poi_fused__to_copy_cat_0
69
+ REDUCE 3.336 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310722048i64i64_tensorptrf32gmemalign16o1310722048i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
70
+ 2.716 ms cudnn_generated_fort_native_sdpa_sm100_flash_fprop_f16_knob_7_128x128x128_4x1x1_kernel0_0
71
+ 1.959 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_TNT
72
+ 1.938 ms kernel_cutlass_kernel_moonepplanningPlanningKernel_object_at__tensorptri32_gmem_align16_o_641_tensorptri32_gmem_align16_o_327681_tensorptri32_gmem_align16_o_41943041_tensorptri32_gmem_ali_0
73
+ 1.918 ms nvjet_qqsss_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NTT
74
+ REDUCE 1.911 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310721024i64i64_tensorptrf32gmemalign16o1310721024i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
75
+ 1.820 ms reduce_kernel
76
+ 1.760 ms CatArrayBatchedCopy_alignedK_contig
77
+ 1.721 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NNT
78
+ 1.717 ms _scatter_gather_elementwise_kernel
79
+ 1.682 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072204811i640_CopyAtom_ThrID10_TVLayoutSrc_0
80
+ 1.550 ms kernel_cutlass_kernel_moonepcombine_prologueCombinePrologueKernel_object_at__tensorptrbf16gmemalign16o348002048i64i64_tensorptri32gmemalign16o1044001_tensorptri32gmemalign16o348001_tensor_0
81
+ 1.459 ms gatherTopK
82
+ 1.328 ms triton_red_fused__to_copy_mul_silu_sum_1
83
+ 1.272 ms fused_rope_forward_kernel
84
+ 1.255 ms multi_tensor_swizzle_row_scaling_kernel
85
+ 1.206 ms setGroupedGemmWgradArguments
86
+ 1.160 ms fused_rope_backward_kernel
87
+ 1.149 ms multi_tensor_swizzle_col_scaling_kernel
88
+ 1.089 ms triton_poi_fused__to_copy_mul_silu_0
89
+ 1.049 ms swizzle_col_scaling_kernel
90
+ 0.973 ms nvjet_tst_256x256_64x4_2x2_2cta_h_bz_NNT
91
+ 0.955 ms nvjet_tss_128x256_64x6_2x1_2cta_v_badd_NTT
92
+ 0.949 ms rmsnorm_bwd_tuned_kernel
93
+ 0.873 ms nvjet_tst_128x256_64x6_2x1_2cta_v_bz_TNT
94
+ 0.856 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072102411i640_CopyAtom_ThrID10_TVLayoutSrc_0
95
+
96
+ === OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.sqlite (device 1, 2 iteration(s)) ===
97
+ reduce_grad kernels : 64 (32/iter)
98
+ reduce_grad sum : 2.624 ms/iter
99
+ reduce_grad busy (union) : 2.624 ms/iter
100
+ ... overlapped with compute: 2.342 ms/iter ( 89.3 %)
101
+ ... exposed on critical path: 0.281 ms/iter ( 10.7 %)
102
+ other GPU work (union) : 99.101 ms/iter
103
+ traced wall : 126.390 ms/iter
104
+ --> time hidden by overlap : 2.342 ms/iter
105
+
106
+ === kernels on device 2, by total time ===
107
+ 26.914 ms elementwise_kernel
108
+ 20.770 ms device_kernel
109
+ 17.491 ms kernel_cutlass_kernel_moonepdispatchDispatchKernel_object_at__tensorptrbf16gmemalign16o40962048i64i64_tensorptrbf16gmemalign16o2785282048i64i64_tensorptri32gmemalign16o327681_tensorptri32_0
110
+ 16.609 ms kernel_cutlass_kernel_moonepcombineCombineKernel_object_at__tensorptrbf16gmemalign16o2785282048i64i64_tensorptrbf16gmemalign16o40962048i64i64_tensorptri32gmemalign16o327681_tensorptri32gm_0
111
+ 14.132 ms multi_tensor_apply_kernel
112
+ 12.993 ms cast_mxfp8_2D_kernel
113
+ 12.962 ms vectorized_elementwise_kernel
114
+ 8.101 ms ncclDevKernel_ReduceScatter_Sum_f32_RING_LL
115
+ 8.067 ms ncclDevKernel_AllReduce_Sum_f32_RING_LL
116
+ 6.240 ms unrolled_elementwise_kernel
117
+ 6.058 ms cudnn_generated_fort_native_sdpa_sm100_flash_bprop_f16_knob_31_128x128x128_1x4x1_kernel0_0
118
+ 4.115 ms kernel_cutlass_kernel_moonepinter_rank_syncInterRankSyncKernel_object_at__tensorptri32gmemalign16o41943041_tensorptri32gmemalign16o10___0
119
+ 3.701 ms triton_poi_fused__to_copy_cat_0
120
+ 3.437 ms ncclDevKernel_AllGather_RING_LL
121
+ REDUCE 3.200 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310722048i64i64_tensorptrf32gmemalign16o1310722048i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
122
+ 2.795 ms cudnn_generated_fort_native_sdpa_sm100_flash_fprop_f16_knob_7_128x128x128_4x1x1_kernel0_0
123
+ 1.975 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_TNT
124
+ 1.951 ms nvjet_qqsss_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NTT
125
+ 1.948 ms kernel_cutlass_kernel_moonepplanningPlanningKernel_object_at__tensorptri32_gmem_align16_o_641_tensorptri32_gmem_align16_o_327681_tensorptri32_gmem_align16_o_41943041_tensorptri32_gmem_ali_0
126
+ REDUCE 1.910 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310721024i64i64_tensorptrf32gmemalign16o1310721024i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
127
+ 1.817 ms reduce_kernel
128
+ 1.759 ms CatArrayBatchedCopy_alignedK_contig
129
+ 1.720 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NNT
130
+ 1.698 ms _scatter_gather_elementwise_kernel
131
+ 1.579 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072204811i640_CopyAtom_ThrID10_TVLayoutSrc_0
132
+ 1.557 ms kernel_cutlass_kernel_moonepcombine_prologueCombinePrologueKernel_object_at__tensorptrbf16gmemalign16o348002048i64i64_tensorptri32gmemalign16o1044001_tensorptri32gmemalign16o348001_tensor_0
133
+ 1.463 ms gatherTopK
134
+ 1.310 ms triton_red_fused__to_copy_mul_silu_sum_1
135
+ 1.260 ms fused_rope_forward_kernel
136
+ 1.235 ms multi_tensor_swizzle_row_scaling_kernel
137
+ 1.207 ms setGroupedGemmWgradArguments
138
+ 1.156 ms fused_rope_backward_kernel
139
+ 1.153 ms multi_tensor_swizzle_col_scaling_kernel
140
+ 1.091 ms triton_poi_fused__to_copy_mul_silu_0
141
+ 1.028 ms swizzle_col_scaling_kernel
142
+ 0.989 ms nvjet_tss_128x256_64x6_2x1_2cta_v_badd_NTT
143
+ 0.975 ms nvjet_tst_256x256_64x4_2x2_2cta_h_bz_NNT
144
+ 0.944 ms rmsnorm_bwd_tuned_kernel
145
+ 0.886 ms nvjet_tst_128x256_64x6_2x1_2cta_v_bz_TNT
146
+ 0.764 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072102411i640_CopyAtom_ThrID10_TVLayoutSrc_0
147
+
148
+ === OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.sqlite (device 2, 2 iteration(s)) ===
149
+ reduce_grad kernels : 64 (32/iter)
150
+ reduce_grad sum : 2.555 ms/iter
151
+ reduce_grad busy (union) : 2.555 ms/iter
152
+ ... overlapped with compute: 2.266 ms/iter ( 88.7 %)
153
+ ... exposed on critical path: 0.289 ms/iter ( 11.3 %)
154
+ other GPU work (union) : 96.330 ms/iter
155
+ traced wall : 126.336 ms/iter
156
+ --> time hidden by overlap : 2.266 ms/iter
157
+
158
+ === kernels on device 3, by total time ===
159
+ 26.933 ms elementwise_kernel
160
+ 21.033 ms device_kernel
161
+ 17.470 ms kernel_cutlass_kernel_moonepdispatchDispatchKernel_object_at__tensorptrbf16gmemalign16o40962048i64i64_tensorptrbf16gmemalign16o2785282048i64i64_tensorptri32gmemalign16o327681_tensorptri32_0
162
+ 16.650 ms kernel_cutlass_kernel_moonepcombineCombineKernel_object_at__tensorptrbf16gmemalign16o2785282048i64i64_tensorptrbf16gmemalign16o40962048i64i64_tensorptri32gmemalign16o327681_tensorptri32gm_0
163
+ 14.120 ms multi_tensor_apply_kernel
164
+ 13.560 ms vectorized_elementwise_kernel
165
+ 13.004 ms cast_mxfp8_2D_kernel
166
+ 10.649 ms ncclDevKernel_AllGather_RING_LL
167
+ 10.631 ms ncclDevKernel_AllReduce_Sum_f32_RING_LL
168
+ 8.082 ms ncclDevKernel_ReduceScatter_Sum_f32_RING_LL
169
+ 6.252 ms unrolled_elementwise_kernel
170
+ 6.172 ms cudnn_generated_fort_native_sdpa_sm100_flash_bprop_f16_knob_31_128x128x128_1x4x1_kernel0_0
171
+ 3.707 ms triton_poi_fused__to_copy_cat_0
172
+ 3.515 ms kernel_cutlass_kernel_moonepinter_rank_syncInterRankSyncKernel_object_at__tensorptri32gmemalign16o41943041_tensorptri32gmemalign16o10___0
173
+ REDUCE 3.216 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310722048i64i64_tensorptrf32gmemalign16o1310722048i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
174
+ 2.866 ms cudnn_generated_fort_native_sdpa_sm100_flash_fprop_f16_knob_7_128x128x128_4x1x1_kernel0_0
175
+ 1.984 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_TNT
176
+ 1.955 ms nvjet_qqsss_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NTT
177
+ 1.955 ms kernel_cutlass_kernel_moonepplanningPlanningKernel_object_at__tensorptri32_gmem_align16_o_641_tensorptri32_gmem_align16_o_327681_tensorptri32_gmem_align16_o_41943041_tensorptri32_gmem_ali_0
178
+ REDUCE 1.907 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310721024i64i64_tensorptrf32gmemalign16o1310721024i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
179
+ 1.820 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072204811i640_CopyAtom_ThrID10_TVLayoutSrc_0
180
+ 1.805 ms reduce_kernel
181
+ 1.759 ms CatArrayBatchedCopy_alignedK_contig
182
+ 1.729 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NNT
183
+ 1.689 ms _scatter_gather_elementwise_kernel
184
+ 1.560 ms kernel_cutlass_kernel_moonepcombine_prologueCombinePrologueKernel_object_at__tensorptrbf16gmemalign16o348002048i64i64_tensorptri32gmemalign16o1044001_tensorptri32gmemalign16o348001_tensor_0
185
+ 1.457 ms gatherTopK
186
+ 1.319 ms triton_red_fused__to_copy_mul_silu_sum_1
187
+ 1.259 ms fused_rope_forward_kernel
188
+ 1.232 ms multi_tensor_swizzle_row_scaling_kernel
189
+ 1.184 ms setGroupedGemmWgradArguments
190
+ 1.156 ms multi_tensor_swizzle_col_scaling_kernel
191
+ 1.153 ms fused_rope_backward_kernel
192
+ 1.089 ms triton_poi_fused__to_copy_mul_silu_0
193
+ 1.033 ms swizzle_col_scaling_kernel
194
+ 0.990 ms nvjet_tst_256x256_64x4_2x2_2cta_h_bz_NNT
195
+ 0.984 ms nvjet_tss_128x256_64x6_2x1_2cta_v_badd_NTT
196
+ 0.931 ms rmsnorm_bwd_tuned_kernel
197
+ 0.913 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072102411i640_CopyAtom_ThrID10_TVLayoutSrc_0
198
+ 0.893 ms nvjet_tst_128x256_64x6_2x1_2cta_v_bz_TNT
199
+
200
+ === OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.sqlite (device 3, 2 iteration(s)) ===
201
+ reduce_grad kernels : 64 (32/iter)
202
+ reduce_grad sum : 2.561 ms/iter
203
+ reduce_grad busy (union) : 2.561 ms/iter
204
+ ... overlapped with compute: 2.271 ms/iter ( 88.7 %)
205
+ ... exposed on critical path: 0.291 ms/iter ( 11.3 %)
206
+ other GPU work (union) : 101.598 ms/iter
207
+ traced wall : 170.641 ms/iter
208
+ --> time hidden by overlap : 2.271 ms/iter
209
+
210
+ === kernels on device 4, by total time ===
211
+ 26.937 ms elementwise_kernel
212
+ 20.940 ms device_kernel
213
+ 17.490 ms kernel_cutlass_kernel_moonepdispatchDispatchKernel_object_at__tensorptrbf16gmemalign16o40962048i64i64_tensorptrbf16gmemalign16o2785282048i64i64_tensorptri32gmemalign16o327681_tensorptri32_0
214
+ 16.666 ms kernel_cutlass_kernel_moonepcombineCombineKernel_object_at__tensorptrbf16gmemalign16o2785282048i64i64_tensorptrbf16gmemalign16o40962048i64i64_tensorptri32gmemalign16o327681_tensorptri32gm_0
215
+ 14.130 ms multi_tensor_apply_kernel
216
+ 13.507 ms vectorized_elementwise_kernel
217
+ 13.030 ms cast_mxfp8_2D_kernel
218
+ 12.090 ms ncclDevKernel_AllReduce_Sum_f32_RING_LL
219
+ 8.169 ms ncclDevKernel_ReduceScatter_Sum_f32_RING_LL
220
+ 6.267 ms unrolled_elementwise_kernel
221
+ 6.063 ms cudnn_generated_fort_native_sdpa_sm100_flash_bprop_f16_knob_31_128x128x128_1x4x1_kernel0_0
222
+ 4.340 ms ncclDevKernel_AllGather_RING_LL
223
+ 3.702 ms triton_poi_fused__to_copy_cat_0
224
+ 3.318 ms kernel_cutlass_kernel_moonepinter_rank_syncInterRankSyncKernel_object_at__tensorptri32gmemalign16o41943041_tensorptri32gmemalign16o10___0
225
+ REDUCE 3.244 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310722048i64i64_tensorptrf32gmemalign16o1310722048i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
226
+ 2.783 ms cudnn_generated_fort_native_sdpa_sm100_flash_fprop_f16_knob_7_128x128x128_4x1x1_kernel0_0
227
+ 1.982 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_TNT
228
+ 1.968 ms nvjet_qqsss_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NTT
229
+ 1.965 ms kernel_cutlass_kernel_moonepplanningPlanningKernel_object_at__tensorptri32_gmem_align16_o_641_tensorptri32_gmem_align16_o_327681_tensorptri32_gmem_align16_o_41943041_tensorptri32_gmem_ali_0
230
+ REDUCE 1.893 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310721024i64i64_tensorptrf32gmemalign16o1310721024i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
231
+ 1.850 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072204811i640_CopyAtom_ThrID10_TVLayoutSrc_0
232
+ 1.808 ms reduce_kernel
233
+ 1.761 ms CatArrayBatchedCopy_alignedK_contig
234
+ 1.737 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NNT
235
+ 1.692 ms _scatter_gather_elementwise_kernel
236
+ 1.542 ms kernel_cutlass_kernel_moonepcombine_prologueCombinePrologueKernel_object_at__tensorptrbf16gmemalign16o348002048i64i64_tensorptri32gmemalign16o1044001_tensorptri32gmemalign16o348001_tensor_0
237
+ 1.473 ms gatherTopK
238
+ 1.327 ms triton_red_fused__to_copy_mul_silu_sum_1
239
+ 1.262 ms fused_rope_forward_kernel
240
+ 1.250 ms multi_tensor_swizzle_row_scaling_kernel
241
+ 1.200 ms setGroupedGemmWgradArguments
242
+ 1.160 ms fused_rope_backward_kernel
243
+ 1.150 ms multi_tensor_swizzle_col_scaling_kernel
244
+ 1.089 ms triton_poi_fused__to_copy_mul_silu_0
245
+ 1.019 ms swizzle_col_scaling_kernel
246
+ 0.987 ms nvjet_tss_128x256_64x6_2x1_2cta_v_badd_NTT
247
+ 0.982 ms nvjet_tst_256x256_64x4_2x2_2cta_h_bz_NNT
248
+ 0.936 ms rmsnorm_bwd_tuned_kernel
249
+ 0.885 ms nvjet_tst_128x256_64x6_2x1_2cta_v_bz_TNT
250
+ 0.866 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072102411i640_CopyAtom_ThrID10_TVLayoutSrc_0
251
+
252
+ === OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.sqlite (device 4, 2 iteration(s)) ===
253
+ reduce_grad kernels : 64 (32/iter)
254
+ reduce_grad sum : 2.568 ms/iter
255
+ reduce_grad busy (union) : 2.568 ms/iter
256
+ ... overlapped with compute: 2.287 ms/iter ( 89.1 %)
257
+ ... exposed on critical path: 0.281 ms/iter ( 10.9 %)
258
+ other GPU work (union) : 99.157 ms/iter
259
+ traced wall : 171.982 ms/iter
260
+ --> time hidden by overlap : 2.287 ms/iter
261
+
262
+ === kernels on device 5, by total time ===
263
+ 26.961 ms elementwise_kernel
264
+ 20.928 ms device_kernel
265
+ 17.468 ms kernel_cutlass_kernel_moonepdispatchDispatchKernel_object_at__tensorptrbf16gmemalign16o40962048i64i64_tensorptrbf16gmemalign16o2785282048i64i64_tensorptri32gmemalign16o327681_tensorptri32_0
266
+ 16.619 ms kernel_cutlass_kernel_moonepcombineCombineKernel_object_at__tensorptrbf16gmemalign16o2785282048i64i64_tensorptrbf16gmemalign16o40962048i64i64_tensorptri32gmemalign16o327681_tensorptri32gm_0
267
+ 14.173 ms multi_tensor_apply_kernel
268
+ 13.017 ms cast_mxfp8_2D_kernel
269
+ 12.998 ms vectorized_elementwise_kernel
270
+ 8.019 ms ncclDevKernel_ReduceScatter_Sum_f32_RING_LL
271
+ 6.267 ms unrolled_elementwise_kernel
272
+ 6.063 ms cudnn_generated_fort_native_sdpa_sm100_flash_bprop_f16_knob_31_128x128x128_1x4x1_kernel0_0
273
+ 4.522 ms ncclDevKernel_AllReduce_Sum_f32_RING_LL
274
+ 4.368 ms ncclDevKernel_AllGather_RING_LL
275
+ 3.709 ms triton_poi_fused__to_copy_cat_0
276
+ 3.567 ms kernel_cutlass_kernel_moonepinter_rank_syncInterRankSyncKernel_object_at__tensorptri32gmemalign16o41943041_tensorptri32gmemalign16o10___0
277
+ REDUCE 3.211 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310722048i64i64_tensorptrf32gmemalign16o1310722048i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
278
+ 2.840 ms cudnn_generated_fort_native_sdpa_sm100_flash_fprop_f16_knob_7_128x128x128_4x1x1_kernel0_0
279
+ 1.971 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_TNT
280
+ 1.961 ms kernel_cutlass_kernel_moonepplanningPlanningKernel_object_at__tensorptri32_gmem_align16_o_641_tensorptri32_gmem_align16_o_327681_tensorptri32_gmem_align16_o_41943041_tensorptri32_gmem_ali_0
281
+ 1.952 ms nvjet_qqsss_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NTT
282
+ REDUCE 1.904 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310721024i64i64_tensorptrf32gmemalign16o1310721024i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
283
+ 1.805 ms reduce_kernel
284
+ 1.761 ms CatArrayBatchedCopy_alignedK_contig
285
+ 1.736 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NNT
286
+ 1.724 ms _scatter_gather_elementwise_kernel
287
+ 1.554 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072204811i640_CopyAtom_ThrID10_TVLayoutSrc_0
288
+ 1.547 ms kernel_cutlass_kernel_moonepcombine_prologueCombinePrologueKernel_object_at__tensorptrbf16gmemalign16o348002048i64i64_tensorptri32gmemalign16o1044001_tensorptri32gmemalign16o348001_tensor_0
289
+ 1.472 ms gatherTopK
290
+ 1.323 ms triton_red_fused__to_copy_mul_silu_sum_1
291
+ 1.266 ms fused_rope_forward_kernel
292
+ 1.235 ms multi_tensor_swizzle_row_scaling_kernel
293
+ 1.179 ms setGroupedGemmWgradArguments
294
+ 1.160 ms fused_rope_backward_kernel
295
+ 1.154 ms multi_tensor_swizzle_col_scaling_kernel
296
+ 1.088 ms triton_poi_fused__to_copy_mul_silu_0
297
+ 1.020 ms swizzle_col_scaling_kernel
298
+ 0.987 ms nvjet_tss_128x256_64x6_2x1_2cta_v_badd_NTT
299
+ 0.975 ms nvjet_tst_256x256_64x4_2x2_2cta_h_bz_NNT
300
+ 0.947 ms rmsnorm_bwd_tuned_kernel
301
+ 0.886 ms nvjet_tst_128x256_64x6_2x1_2cta_v_bz_TNT
302
+ 0.846 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072102411i640_CopyAtom_ThrID10_TVLayoutSrc_0
303
+
304
+ === OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.sqlite (device 5, 2 iteration(s)) ===
305
+ reduce_grad kernels : 64 (32/iter)
306
+ reduce_grad sum : 2.558 ms/iter
307
+ reduce_grad busy (union) : 2.558 ms/iter
308
+ ... overlapped with compute: 2.278 ms/iter ( 89.1 %)
309
+ ... exposed on critical path: 0.279 ms/iter ( 10.9 %)
310
+ other GPU work (union) : 95.086 ms/iter
311
+ traced wall : 126.432 ms/iter
312
+ --> time hidden by overlap : 2.278 ms/iter
313
+
314
+ === kernels on device 6, by total time ===
315
+ 26.973 ms elementwise_kernel
316
+ 21.078 ms device_kernel
317
+ 17.446 ms kernel_cutlass_kernel_moonepdispatchDispatchKernel_object_at__tensorptrbf16gmemalign16o40962048i64i64_tensorptrbf16gmemalign16o2785282048i64i64_tensorptri32gmemalign16o327681_tensorptri32_0
318
+ 16.391 ms kernel_cutlass_kernel_moonepcombineCombineKernel_object_at__tensorptrbf16gmemalign16o2785282048i64i64_tensorptrbf16gmemalign16o40962048i64i64_tensorptri32gmemalign16o327681_tensorptri32gm_0
319
+ 14.110 ms multi_tensor_apply_kernel
320
+ 13.644 ms vectorized_elementwise_kernel
321
+ 13.056 ms cast_mxfp8_2D_kernel
322
+ 7.956 ms ncclDevKernel_ReduceScatter_Sum_f32_RING_LL
323
+ 6.887 ms ncclDevKernel_AllReduce_Sum_f32_RING_LL
324
+ 6.290 ms unrolled_elementwise_kernel
325
+ 6.183 ms cudnn_generated_fort_native_sdpa_sm100_flash_bprop_f16_knob_31_128x128x128_1x4x1_kernel0_0
326
+ 4.237 ms ncclDevKernel_AllGather_RING_LL
327
+ 3.708 ms kernel_cutlass_kernel_moonepinter_rank_syncInterRankSyncKernel_object_at__tensorptri32gmemalign16o41943041_tensorptri32gmemalign16o10___0
328
+ 3.706 ms triton_poi_fused__to_copy_cat_0
329
+ REDUCE 3.393 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310722048i64i64_tensorptrf32gmemalign16o1310722048i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
330
+ 2.762 ms cudnn_generated_fort_native_sdpa_sm100_flash_fprop_f16_knob_7_128x128x128_4x1x1_kernel0_0
331
+ 1.988 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_TNT
332
+ 1.959 ms nvjet_qqsss_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NTT
333
+ 1.943 ms kernel_cutlass_kernel_moonepplanningPlanningKernel_object_at__tensorptri32_gmem_align16_o_641_tensorptri32_gmem_align16_o_327681_tensorptri32_gmem_align16_o_41943041_tensorptri32_gmem_ali_0
334
+ REDUCE 1.873 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310721024i64i64_tensorptrf32gmemalign16o1310721024i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
335
+ 1.818 ms reduce_kernel
336
+ 1.765 ms CatArrayBatchedCopy_alignedK_contig
337
+ 1.758 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NNT
338
+ 1.725 ms _scatter_gather_elementwise_kernel
339
+ 1.704 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072204811i640_CopyAtom_ThrID10_TVLayoutSrc_0
340
+ 1.491 ms kernel_cutlass_kernel_moonepcombine_prologueCombinePrologueKernel_object_at__tensorptrbf16gmemalign16o348002048i64i64_tensorptri32gmemalign16o1044001_tensorptri32gmemalign16o348001_tensor_0
341
+ 1.474 ms gatherTopK
342
+ 1.324 ms triton_red_fused__to_copy_mul_silu_sum_1
343
+ 1.263 ms fused_rope_forward_kernel
344
+ 1.227 ms multi_tensor_swizzle_row_scaling_kernel
345
+ 1.191 ms setGroupedGemmWgradArguments
346
+ 1.161 ms fused_rope_backward_kernel
347
+ 1.154 ms multi_tensor_swizzle_col_scaling_kernel
348
+ 1.088 ms triton_poi_fused__to_copy_mul_silu_0
349
+ 1.046 ms swizzle_col_scaling_kernel
350
+ 1.000 ms nvjet_tss_128x256_64x6_2x1_2cta_v_badd_NTT
351
+ 0.986 ms nvjet_tst_256x256_64x4_2x2_2cta_h_bz_NNT
352
+ 0.954 ms rmsnorm_bwd_tuned_kernel
353
+ 0.895 ms nvjet_tst_128x256_64x6_2x1_2cta_v_bz_TNT
354
+ 0.857 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072102411i640_CopyAtom_ThrID10_TVLayoutSrc_0
355
+
356
+ === OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.sqlite (device 6, 2 iteration(s)) ===
357
+ reduce_grad kernels : 64 (32/iter)
358
+ reduce_grad sum : 2.633 ms/iter
359
+ reduce_grad busy (union) : 2.633 ms/iter
360
+ ... overlapped with compute: 2.351 ms/iter ( 89.3 %)
361
+ ... exposed on critical path: 0.282 ms/iter ( 10.7 %)
362
+ other GPU work (union) : 96.552 ms/iter
363
+ traced wall : 174.692 ms/iter
364
+ --> time hidden by overlap : 2.351 ms/iter
365
+
366
+ === kernels on device 7, by total time ===
367
+ 26.961 ms elementwise_kernel
368
+ 20.879 ms device_kernel
369
+ 17.459 ms kernel_cutlass_kernel_moonepdispatchDispatchKernel_object_at__tensorptrbf16gmemalign16o40962048i64i64_tensorptrbf16gmemalign16o2785282048i64i64_tensorptri32gmemalign16o327681_tensorptri32_0
370
+ 15.927 ms kernel_cutlass_kernel_moonepcombineCombineKernel_object_at__tensorptrbf16gmemalign16o2785282048i64i64_tensorptrbf16gmemalign16o40962048i64i64_tensorptri32gmemalign16o327681_tensorptri32gm_0
371
+ 14.149 ms multi_tensor_apply_kernel
372
+ 13.655 ms vectorized_elementwise_kernel
373
+ 13.136 ms ncclDevKernel_AllReduce_Sum_f32_RING_LL
374
+ 13.043 ms cast_mxfp8_2D_kernel
375
+ 8.097 ms ncclDevKernel_ReduceScatter_Sum_f32_RING_LL
376
+ 6.324 ms unrolled_elementwise_kernel
377
+ 6.082 ms cudnn_generated_fort_native_sdpa_sm100_flash_bprop_f16_knob_31_128x128x128_1x4x1_kernel0_0
378
+ 4.182 ms kernel_cutlass_kernel_moonepinter_rank_syncInterRankSyncKernel_object_at__tensorptri32gmemalign16o41943041_tensorptri32gmemalign16o10___0
379
+ 4.152 ms ncclDevKernel_AllGather_RING_LL
380
+ 3.696 ms triton_poi_fused__to_copy_cat_0
381
+ REDUCE 3.560 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310722048i64i64_tensorptrf32gmemalign16o1310722048i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
382
+ 2.737 ms cudnn_generated_fort_native_sdpa_sm100_flash_fprop_f16_knob_7_128x128x128_4x1x1_kernel0_0
383
+ 1.970 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_TNT
384
+ 1.960 ms kernel_cutlass_kernel_moonepplanningPlanningKernel_object_at__tensorptri32_gmem_align16_o_641_tensorptri32_gmem_align16_o_327681_tensorptri32_gmem_align16_o_41943041_tensorptri32_gmem_ali_0
385
+ 1.938 ms nvjet_qqsss_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NTT
386
+ REDUCE 1.881 ms kernel_cutlass_kernel_moonepgrad_reduceGradReduceKernel_object_at__tensorptrf32gmemalign16o1310721024i64i64_tensorptrf32gmemalign16o1310721024i64i64_CopyAtom_ThrID10_TVLayoutSrc11638401_T_0
387
+ 1.847 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072204811i640_CopyAtom_ThrID10_TVLayoutSrc_0
388
+ 1.808 ms reduce_kernel
389
+ 1.762 ms CatArrayBatchedCopy_alignedK_contig
390
+ 1.713 ms nvjet_qqtst_128x256_128x6_2x2_2cta_h_bz_Avec32UE8M0_Bvec32UE8M0_NNT
391
+ 1.704 ms _scatter_gather_elementwise_kernel
392
+ 1.634 ms kernel_cutlass_kernel_moonepcombine_prologueCombinePrologueKernel_object_at__tensorptrbf16gmemalign16o348002048i64i64_tensorptri32gmemalign16o1044001_tensorptri32gmemalign16o348001_tensor_0
393
+ 1.457 ms gatherTopK
394
+ 1.325 ms triton_red_fused__to_copy_mul_silu_sum_1
395
+ 1.254 ms fused_rope_forward_kernel
396
+ 1.250 ms multi_tensor_swizzle_row_scaling_kernel
397
+ 1.186 ms setGroupedGemmWgradArguments
398
+ 1.159 ms fused_rope_backward_kernel
399
+ 1.154 ms multi_tensor_swizzle_col_scaling_kernel
400
+ 1.088 ms triton_poi_fused__to_copy_mul_silu_0
401
+ 1.056 ms swizzle_col_scaling_kernel
402
+ 0.985 ms nvjet_tst_256x256_64x4_2x2_2cta_h_bz_NNT
403
+ 0.980 ms nvjet_tss_128x256_64x6_2x1_2cta_v_badd_NTT
404
+ 0.947 ms rmsnorm_bwd_tuned_kernel
405
+ 0.880 ms nvjet_tst_128x256_64x6_2x1_2cta_v_bz_TNT
406
+ 0.872 ms kernel_cutlass_kernel_moonepprefetchPrefetchKernel_object_at__CopyAtom_ThrID10_TVLayoutSrc11638401_TVLayoutDst11638401_Valuetypebf16_tensor00o131072102411i640_CopyAtom_ThrID10_TVLayoutSrc_0
407
+
408
+ === OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.sqlite (device 7, 2 iteration(s)) ===
409
+ reduce_grad kernels : 64 (32/iter)
410
+ reduce_grad sum : 2.720 ms/iter
411
+ reduce_grad busy (union) : 2.720 ms/iter
412
+ ... overlapped with compute: 2.444 ms/iter ( 89.8 %)
413
+ ... exposed on critical path: 0.277 ms/iter ( 10.2 %)
414
+ other GPU work (union) : 99.620 ms/iter
415
+ traced wall : 172.057 ms/iter
416
+ --> time hidden by overlap : 2.444 ms/iter
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/moonep_async_overlap_summary.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ devices=8 reduce=2.613 ms/iter overlap=2.330 ms/iter (89.2%) exposed=0.283 ms/iter status=PASS
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/run.env ADDED
@@ -0,0 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ backend=moonep
2
+ config=OLMOE_1B_7B_L16
3
+ family=tierB
4
+ source_model=OLMoE-1B-7B-0924
5
+ standard_num_layers=16
6
+ base_rung=
7
+ repo=/workspace/Megatron-MoE-Echo
8
+ branch=moonep_backend_async_grad_reduce
9
+ commit=2e68c8ad
10
+ moonep_source_dir=/workspace/MoonEP
11
+ moonep_source_branch=megatron-ddp-symmetric-params
12
+ moonep_source_commit=d8bc666
13
+ num_layers=16
14
+ hidden=2048
15
+ ffn_hidden=1024
16
+ moe_ffn_hidden=1024
17
+ num_experts=64
18
+ topk=8
19
+ experts_per_rank=8
20
+ expert_params=6291456
21
+ heads=16
22
+ kv_groups=16
23
+ kv_channels=128
24
+ qk_layernorm=0
25
+ shared_ffn=0
26
+ shared_expert_overlap=0
27
+ moe_token_dispatcher=flex
28
+ mla=0
29
+ q_lora_rank=
30
+ kv_lora_rank=512
31
+ qk_head_dim=128
32
+ qk_pos_emb_head_dim=64
33
+ v_head_dim=128
34
+ dense_first_layers=0
35
+ moe_layers=16
36
+ rotary_base=10000
37
+ rotary_percent=1.0
38
+ rotary_scaling=
39
+ mscale=
40
+ mscale_all_dim=
41
+ qkv_bias=0
42
+ router_balancing=aux_loss
43
+ aux_loss_coeff=1e-2
44
+ router_pre_softmax=1
45
+ router_fusion=0
46
+ router_dtype=fp32
47
+ router_groups=
48
+ router_group_topk=
49
+ router_score_function=softmax
50
+ router_expert_bias=0
51
+ router_bias_update_rate=
52
+ router_topk_scaling_factor=
53
+ shared_expert_gate=0
54
+ apply_layernorm_1p=0
55
+ attention_output_gate=0
56
+ linear_attention_type=
57
+ linear_attention_freq=
58
+ linear_conv_kernel_dim=
59
+ linear_key_head_dim=
60
+ linear_value_head_dim=
61
+ linear_num_key_heads=
62
+ linear_num_value_heads=
63
+ mtp_num_layers=0
64
+ mtp_loss_scaling_factor=0.1
65
+ no_weight_decay_cond_type=
66
+ max_position_embeddings=4096
67
+ init_method_std=0.02
68
+ device_initiated_grouped_gemm=1
69
+ norm_epsilon=1e-5
70
+ seq_length=4096
71
+ micro_batch_size=1
72
+ global_batch_size=8
73
+ num_microbatches=1
74
+ assignments_per_rank=32768
75
+ mean_tokens_per_expert=4096
76
+ slot_policy=matched
77
+ slots_per_rank=8
78
+ capacity_mode=none
79
+ capacity=none
80
+ cuda_graph=1
81
+ probe_mode=0
82
+ async_grad_reduce=1
83
+ branch_role=timing
84
+ memory_dump_mode=none
85
+ memory_snapshot_path=
86
+ cuda_graph_warmup_steps=3
87
+ recompute=none
88
+ force_balance=0
89
+ train_iters=10
90
+ ep_size=8
91
+ echo_dispatch_overlap=1
92
+ echo_dispatch_recompute=0
93
+ ultraep_quota_min_tokens=
94
+ ultraep_weight_sync_join=
95
+ profile_enable=1
96
+ kernel_trace=1
97
+ fp8_mode=mxfp8
98
+ fp8_recipe=mxfp8
99
+ training_seed=1234
100
+ learning_rate=1e-4
101
+ min_learning_rate=1.0e-5
102
+ lr_warmup_iters=5
103
+ state_dump_enable=0
104
+ state_dump_dir=
105
+ moonep_assert_no_pending_reduce=1
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/train.log ADDED
@@ -0,0 +1,1139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ==============================================================
2
+ OLMOE_1B_7B_L16 / moonep (slots/rank=8 policy=matched)
3
+ E=64 K=8 H'=1024 m=8
4
+ L=16 seq=4096 gbs=8 (1 microbatches)
5
+ cap_mode=none cap=none cg=1 recompute=none
6
+ memory_dump=none snapshots=/workspace/granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/memory_snapshots/snapshot.pickle
7
+ repo=/workspace/Megatron-MoE-Echo @ 2e68c8ad
8
+ ==============================================================
9
+ W0921 09:09:32.562000 806907 torch/distributed/run.py:766]
10
+ W0921 09:09:32.562000 806907 torch/distributed/run.py:766] *****************************************
11
+ W0921 09:09:32.562000 806907 torch/distributed/run.py:766] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
12
+ W0921 09:09:32.562000 806907 torch/distributed/run.py:766] *****************************************
13
+ using world size: 8, data-parallel size: 8, context-parallel size: 1, hierarchical context-parallel sizes: None, tensor-model-parallel size: 1, pipeline-model-parallel size: 1
14
+ WARNING: overriding default arguments for tokenizer_type:GPT2BPETokenizer with tokenizer_type:GPT2BPETokenizer
15
+ Number of virtual stages per pipeline stage: None
16
+ accumulate and all-reduce gradients in fp32 for bfloat16 data type.
17
+ using torch.bfloat16 for parameters ...
18
+ /workspace/Megatron-MoE-Echo/megatron/training/arguments.py:831: UserWarning: Disabling sequence parallelism because tensor model parallelism is disabled
19
+ warnings.warn("Disabling sequence parallelism because tensor model parallelism is disabled")
20
+ ------------------------ arguments ------------------------
21
+ account_for_embedding_in_pipeline_split ......... False
22
+ account_for_loss_in_pipeline_split .............. False
23
+ accumulate_allreduce_grads_in_fp32 .............. True
24
+ adam_beta1 ...................................... 0.9
25
+ adam_beta2 ...................................... 0.999
26
+ adam_eps ........................................ 1e-08
27
+ add_bias_linear ................................. False
28
+ add_position_embedding .......................... False
29
+ add_qkv_bias .................................... False
30
+ adlr_autoresume ................................. False
31
+ adlr_autoresume_interval ........................ 1000
32
+ align_grad_reduce ............................... True
33
+ align_param_gather .............................. False
34
+ app_tag_run_name ................................ None
35
+ app_tag_run_version ............................. 0.0.0
36
+ apply_layernorm_1p .............................. False
37
+ apply_query_key_layer_scaling ................... False
38
+ apply_residual_connection_post_layernorm ........ False
39
+ apply_rope_fusion ............................... True
40
+ async_save ...................................... None
41
+ async_tensor_model_parallel_allreduce ........... True
42
+ attention_backend ............................... AttnBackend.auto
43
+ attention_dropout ............................... 0.0
44
+ attention_softmax_in_fp32 ....................... False
45
+ auto_detect_ckpt_format ......................... False
46
+ barrier_with_L1_time ............................ True
47
+ bert_binary_head ................................ True
48
+ bert_embedder_type .............................. megatron
49
+ bert_load ....................................... None
50
+ bf16 ............................................ True
51
+ bias_dropout_fusion ............................. True
52
+ bias_gelu_fusion ................................ False
53
+ bias_swiglu_fusion .............................. True
54
+ biencoder_projection_dim ........................ 0
55
+ biencoder_shared_query_context_model ............ False
56
+ block_data_path ................................. None
57
+ calc_ft_timeouts ................................ False
58
+ calculate_per_token_loss ........................ False
59
+ check_for_large_grads ........................... False
60
+ check_for_nan_in_loss_and_grad .................. False
61
+ check_for_spiky_loss ............................ False
62
+ check_weight_hash_across_dp_replicas_interval ... None
63
+ ckpt_assume_constant_structure .................. False
64
+ ckpt_convert_format ............................. None
65
+ ckpt_convert_save ............................... None
66
+ ckpt_convert_update_legacy_dist_opt_format ...... False
67
+ ckpt_format ..................................... torch_dist
68
+ ckpt_fully_parallel_load ........................ False
69
+ ckpt_fully_parallel_save ........................ True
70
+ ckpt_fully_parallel_save_deprecated ............. False
71
+ ckpt_step ....................................... None
72
+ classes_fraction ................................ 1.0
73
+ clip_grad ....................................... 1.0
74
+ clone_scatter_output_in_embedding ............... True
75
+ config_logger_dir ...............................
76
+ consumed_train_samples .......................... 0
77
+ consumed_valid_samples .......................... 0
78
+ context_parallel_size ........................... 1
79
+ cp_comm_type .................................... ['p2p']
80
+ create_attention_mask_in_dataloader ............. True
81
+ cross_entropy_fusion_impl ....................... native
82
+ cross_entropy_loss_fusion ....................... False
83
+ cuda_graph_scope ................................ full_iteration
84
+ cuda_graph_warmup_steps ......................... 3
85
+ data_args_path .................................. None
86
+ data_cache_path ................................. None
87
+ data_parallel_random_init ....................... False
88
+ data_parallel_sharding_strategy ................. no_shard
89
+ data_parallel_size .............................. 8
90
+ data_path ....................................... ['/workspace/data/wikitext/wikitext103_gpt2_text_document']
91
+ data_per_class_fraction ......................... 1.0
92
+ data_sharding ................................... True
93
+ dataloader_type ................................. single
94
+ ddp_average_in_collective ....................... False
95
+ ddp_bucket_size ................................. None
96
+ ddp_num_buckets ................................. None
97
+ ddp_pad_buckets_for_high_nccl_busbw ............. False
98
+ decoder_first_pipeline_num_layers ............... None
99
+ decoder_last_pipeline_num_layers ................ None
100
+ decoder_num_layers .............................. None
101
+ decoder_seq_length .............................. None
102
+ decoupled_lr .................................... None
103
+ decoupled_min_lr ................................ None
104
+ decrease_batch_size_if_needed ................... False
105
+ defer_embedding_wgrad_compute ................... False
106
+ delay_wgrad_compute ............................. False
107
+ deprecated_use_mcore_models ..................... True
108
+ deterministic_mode .............................. False
109
+ dino_bottleneck_size ............................ 256
110
+ dino_freeze_last_layer .......................... 1
111
+ dino_head_hidden_size ........................... 2048
112
+ dino_local_crops_number ......................... 10
113
+ dino_local_img_size ............................. 96
114
+ dino_norm_last_layer ............................ False
115
+ dino_teacher_temp ............................... 0.07
116
+ dino_warmup_teacher_temp ........................ 0.04
117
+ dino_warmup_teacher_temp_epochs ................. 30
118
+ disable_bf16_reduced_precision_matmul ........... False
119
+ disable_mamba_mem_eff_path ...................... False
120
+ disable_straggler_on_startup .................... False
121
+ dist_ckpt_format_deprecated ..................... None
122
+ dist_ckpt_strictness ............................ assume_ok_unexpected
123
+ distribute_saved_activations .................... False
124
+ distributed_backend ............................. nccl
125
+ distributed_timeout_minutes ..................... 10
126
+ embedding_init_method_std ....................... None
127
+ embedding_path .................................. None
128
+ empty_unused_memory_level ....................... 0
129
+ enable_cuda_graph ............................... True
130
+ enable_experimental ............................. False
131
+ enable_ft_package ............................... False
132
+ enable_gloo_process_groups ...................... True
133
+ enable_msc ...................................... True
134
+ enable_one_logger ............................... True
135
+ encoder_num_layers .............................. 16
136
+ encoder_seq_length .............................. 4096
137
+ end_weight_decay ................................ 0.1
138
+ eod_mask_loss ................................... False
139
+ error_injection_rate ............................ 0
140
+ error_injection_type ............................ transient_error
141
+ eval_interval ................................... 1000000
142
+ eval_iters ...................................... 0
143
+ evidence_data_path .............................. None
144
+ exit_duration_in_mins ........................... None
145
+ exit_interval ................................... None
146
+ exit_on_missing_checkpoint ...................... False
147
+ exit_signal_handler ............................. False
148
+ exp_avg_dtype ................................... torch.float32
149
+ exp_avg_sq_dtype ................................ torch.float32
150
+ expert_model_parallel_size ...................... 8
151
+ expert_tensor_parallel_size ..................... 1
152
+ export_force_local_attention .................... False
153
+ export_kd_cfg ................................... None
154
+ export_kd_teacher_ckpt_format ................... None
155
+ export_kd_teacher_load .......................... None
156
+ export_kv_cache_quant ........................... False
157
+ export_legacy_megatron .......................... False
158
+ export_model_type ............................... GPTModel
159
+ export_moe_apply_probs_on_input ................. False
160
+ export_qk_l2_norm ............................... False
161
+ export_quant_cfg ................................ None
162
+ export_real_quant_cfg ........................... None
163
+ export_te_mcore_model ........................... False
164
+ external_cuda_graph ............................. False
165
+ ffn_hidden_size ................................. 1024
166
+ finetune ........................................ False
167
+ finetune_data_split ............................. train
168
+ finetune_hf_dataset ............................. None
169
+ first_last_layers_bf16 .......................... False
170
+ flash_decode .................................... False
171
+ fp16 ............................................ False
172
+ fp16_lm_cross_entropy ........................... False
173
+ fp32_residual_connection ........................ False
174
+ fp8 ............................................. e4m3
175
+ fp8_amax_compute_algo ........................... most_recent
176
+ fp8_amax_history_len ............................ 1
177
+ fp8_interval .................................... 1
178
+ fp8_margin ...................................... 0
179
+ fp8_param_gather ................................ False
180
+ fp8_recipe ...................................... mxfp8
181
+ fp8_wgrad ....................................... True
182
+ fsdp_double_buffer .............................. False
183
+ global_batch_size ............................... 8
184
+ grad_reduce_in_bf16 ............................. False
185
+ gradient_accumulation_fusion .................... True
186
+ gradient_reduce_div_fusion ...................... True
187
+ group_query_attention ........................... True
188
+ head_lr_mult .................................... 1.0
189
+ heterogeneous_layers_config_encoded_json ........ None
190
+ heterogeneous_layers_config_path ................ None
191
+ hidden_dropout .................................. 0.0
192
+ hidden_size ..................................... 2048
193
+ hierarchical_context_parallel_sizes ............. None
194
+ high_priority_stream_groups ..................... []
195
+ hybrid_attention_ratio .......................... 0.0
196
+ hybrid_mlp_ratio ................................ 0.0
197
+ hybrid_override_pattern ......................... None
198
+ hysteresis ...................................... 2
199
+ ict_head_size ................................... None
200
+ ict_load ........................................ None
201
+ img_h ........................................... 224
202
+ img_w ........................................... 224
203
+ indexer_batch_size .............................. 128
204
+ indexer_log_interval ............................ 1000
205
+ inference_batch_times_seqlen_threshold .......... -1
206
+ inference_dynamic_batching ...................... False
207
+ inference_dynamic_batching_buffer_guaranteed_fraction 0.2
208
+ inference_dynamic_batching_buffer_overflow_factor None
209
+ inference_dynamic_batching_buffer_size_gb ....... 40.0
210
+ inference_dynamic_batching_chunk_size ........... 256
211
+ inference_dynamic_batching_max_requests_override None
212
+ inference_dynamic_batching_max_tokens_override .. None
213
+ inference_dynamic_batching_num_cuda_graphs ...... 16
214
+ inference_max_batch_size ........................ 8
215
+ inference_max_seq_length ........................ 2560
216
+ inference_rng_tracker ........................... False
217
+ init_method_std ................................. 0.02
218
+ init_method_xavier_uniform ...................... False
219
+ init_model_with_meta_device ..................... False
220
+ initial_loss_scale .............................. 4294967296
221
+ inprocess_active_world_size ..................... 8
222
+ inprocess_barrier_timeout ....................... 120
223
+ inprocess_completion_timeout .................... 120
224
+ inprocess_empty_cuda_cache ...................... False
225
+ inprocess_granularity ........................... node
226
+ inprocess_hard_timeout .......................... 90
227
+ inprocess_heartbeat_interval .................... 30
228
+ inprocess_heartbeat_timeout ..................... 60
229
+ inprocess_last_call_wait ........................ 1
230
+ inprocess_max_iterations ........................ None
231
+ inprocess_monitor_process_interval .............. 1.0
232
+ inprocess_monitor_thread_interval ............... 1.0
233
+ inprocess_progress_watchdog_interval ............ 1.0
234
+ inprocess_restart ............................... False
235
+ inprocess_soft_timeout .......................... 60
236
+ inprocess_termination_grace_time ................ 1
237
+ is_hybrid_model ................................. False
238
+ iter_per_epoch .................................. 1250
239
+ iterations_to_skip .............................. []
240
+ keep_fp8_transpose_cache_when_using_custom_fsdp . False
241
+ kitchen_config_file ............................. None
242
+ kitchen_recipe_number ........................... None
243
+ kv_channels ..................................... 128
244
+ kv_lora_rank .................................... 32
245
+ lazy_mpu_init ................................... None
246
+ load ............................................ None
247
+ load_main_params_from_ckpt ...................... None
248
+ load_model_opt_format ........................... False
249
+ local_rank ...................................... 0
250
+ log_energy ...................................... False
251
+ log_interval .................................... 1
252
+ log_loss_scale_to_tensorboard ................... True
253
+ log_memory_to_tensorboard ....................... False
254
+ log_num_zeros_in_grad ........................... False
255
+ log_params_norm ................................. False
256
+ log_progress .................................... False
257
+ log_straggler ................................... False
258
+ log_throughput .................................. True
259
+ log_timers_to_tensorboard ....................... False
260
+ log_validation_ppl_to_tensorboard ............... False
261
+ log_world_size_to_tensorboard ................... False
262
+ logging_level ................................... None
263
+ loss_scale ...................................... None
264
+ loss_scale_window ............................... 1000
265
+ lr .............................................. 0.0001
266
+ lr_decay_iters .................................. 8
267
+ lr_decay_samples ................................ None
268
+ lr_decay_style .................................. cosine
269
+ lr_warmup_fraction .............................. None
270
+ lr_warmup_init .................................. 0.0
271
+ lr_warmup_iters ................................. 5
272
+ lr_warmup_samples ............................... 0
273
+ lr_wsd_decay_iters .............................. None
274
+ lr_wsd_decay_samples ............................ None
275
+ lr_wsd_decay_style .............................. exponential
276
+ main_grads_dtype ................................ torch.float32
277
+ main_params_dtype ............................... torch.float32
278
+ make_vocab_size_divisible_by .................... 128
279
+ mamba_head_dim .................................. 64
280
+ mamba_num_groups ................................ 8
281
+ mamba_num_heads ................................. None
282
+ mamba_state_dim ................................. 128
283
+ manual_gc ....................................... False
284
+ manual_gc_eval .................................. True
285
+ manual_gc_interval .............................. 0
286
+ mask_factor ..................................... 1.0
287
+ mask_prob ....................................... 0.15
288
+ mask_type ....................................... random
289
+ masked_softmax_fusion ........................... False
290
+ max_position_embeddings ......................... 4096
291
+ max_tokens_to_oom ............................... 12000
292
+ memory_snapshot_path ............................ snapshot.pickle
293
+ merge_file ...................................... /workspace/data/wikitext/gpt2-merges.txt
294
+ micro_batch_size ................................ 1
295
+ microbatch_group_size_per_vp_stage .............. None
296
+ mid_level_dataset_surplus ....................... 0.005
297
+ min_loss_scale .................................. 1.0
298
+ min_lr .......................................... 1e-05
299
+ mlp_chunks_for_prefill .......................... 1
300
+ mmap_bin_files .................................. True
301
+ mock_data ....................................... False
302
+ moe_apply_probs_on_input ........................ False
303
+ moe_aux_loss_coeff .............................. 0.01
304
+ moe_deepep_num_sms .............................. 20
305
+ moe_echo_enable_random_offloading ............... False
306
+ moe_echo_expert_dispatch_overlap ................ False
307
+ moe_echo_expert_dispatcher_type ................. hybridep
308
+ moe_echo_recompute_expert_dispatch .............. False
309
+ moe_enable_deepep ............................... False
310
+ moe_enable_echo ................................. False
311
+ moe_enable_hybridep ............................. False
312
+ moe_enable_moonep ............................... True
313
+ moe_expert_capacity_factor ...................... None
314
+ moe_extended_tp ................................. False
315
+ moe_ffn_hidden_size ............................. 1024
316
+ moe_grouped_gemm ................................ True
317
+ moe_hybridep_num_sms ............................ 16
318
+ moe_input_jitter_eps ............................ None
319
+ moe_layer_freq .................................. 1
320
+ moe_layer_recompute ............................. False
321
+ moe_moonep_async_grad_reduce .................... True
322
+ moe_moonep_num_sms .............................. None
323
+ moe_num_echo_experts ............................ None
324
+ moe_pad_expert_input_to_capacity ................ False
325
+ moe_per_layer_logging ........................... False
326
+ moe_permute_fusion .............................. False
327
+ moe_received_token_capacity ..................... None
328
+ moe_router_bias_update_rate ..................... 0.001
329
+ moe_router_dtype ................................ fp32
330
+ moe_router_enable_expert_bias ................... False
331
+ moe_router_force_load_balancing ................. False
332
+ moe_router_group_topk ........................... None
333
+ moe_router_load_balancing_type .................. aux_loss
334
+ moe_router_num_groups ........................... None
335
+ moe_router_padding_for_fp8 ...................... False
336
+ moe_router_pre_softmax .......................... True
337
+ moe_router_score_function ....................... softmax
338
+ moe_router_topk ................................. 8
339
+ moe_router_topk_scaling_factor .................. None
340
+ moe_shared_expert_intermediate_size ............. None
341
+ moe_shared_expert_overlap ....................... False
342
+ moe_token_dispatcher_type ....................... flex
343
+ moe_token_drop_policy ........................... probs
344
+ moe_upcycling_granularity ....................... 1
345
+ moe_use_device_initiated_grouped_gemm ........... True
346
+ moe_use_legacy_grouped_gemm ..................... False
347
+ moe_use_upcycling ............................... False
348
+ moe_z_loss_coeff ................................ None
349
+ mrope_section ................................... None
350
+ mscale .......................................... 1.0
351
+ mscale_all_dim .................................. 0.0
352
+ mtp_loss_scaling_factor ......................... 0.1
353
+ mtp_num_layers .................................. None
354
+ multi_latent_attention .......................... False
355
+ nccl_all_reduce_for_prefill ..................... False
356
+ nccl_communicator_config_path ................... None
357
+ nccl_ub ......................................... False
358
+ no_load_optim ................................... None
359
+ no_load_rng ..................................... None
360
+ no_persist_layer_norm ........................... False
361
+ no_rope_freq .................................... None
362
+ no_save_optim ................................... None
363
+ no_save_rng ..................................... None
364
+ non_persistent_ckpt_type ........................ None
365
+ non_persistent_global_ckpt_dir .................. None
366
+ non_persistent_local_ckpt_algo .................. fully_parallel
367
+ non_persistent_local_ckpt_dir ................... None
368
+ non_persistent_save_interval .................... None
369
+ norm_epsilon .................................... 1e-05
370
+ normalization ................................... RMSNorm
371
+ num_attention_heads ............................. 16
372
+ num_channels .................................... 3
373
+ num_classes ..................................... 1000
374
+ num_dataset_builder_threads ..................... 1
375
+ num_distributed_optimizer_instances ............. 1
376
+ num_experts ..................................... 64
377
+ num_layers ...................................... 16
378
+ num_layers_at_end_in_bf16 ....................... 1
379
+ num_layers_at_start_in_bf16 ..................... 1
380
+ num_layers_per_virtual_pipeline_stage ........... None
381
+ num_query_groups ................................ 16
382
+ num_virtual_stages_per_pipeline_rank ............ None
383
+ num_workers ..................................... 2
384
+ object_storage_cache_path ....................... None
385
+ one_logger_async ................................ False
386
+ one_logger_project .............................. megatron-lm
387
+ one_logger_run_name ............................. None
388
+ onnx_safe ....................................... None
389
+ openai_gelu ..................................... False
390
+ optimizer ....................................... adam
391
+ optimizer_cpu_offload ........................... False
392
+ optimizer_offload_fraction ...................... 1.0
393
+ output_bert_embeddings .......................... False
394
+ overlap_cpu_optimizer_d2h_h2d ................... False
395
+ overlap_grad_reduce ............................. True
396
+ overlap_moe_expert_parallel_comm ................ False
397
+ overlap_p2p_comm ................................ False
398
+ overlap_p2p_comm_warmup_flush ................... False
399
+ overlap_param_gather ............................ True
400
+ overlap_param_gather_with_optimizer_step ........ False
401
+ override_opt_param_scheduler .................... False
402
+ params_dtype .................................... torch.bfloat16
403
+ patch_dim ....................................... 16
404
+ per_split_data_args_path ........................ None
405
+ perform_initialization .......................... True
406
+ pin_cpu_grads ................................... True
407
+ pin_cpu_params .................................. True
408
+ pipeline_model_parallel_comm_backend ............ None
409
+ pipeline_model_parallel_layout .................. None
410
+ pipeline_model_parallel_size .................... 1
411
+ position_embedding_type ......................... rope
412
+ pretrained_checkpoint ........................... None
413
+ profile ......................................... True
414
+ profile_ranks ................................... [0, 1, 2, 3, 4, 5, 6, 7]
415
+ profile_step_end ................................ 8
416
+ profile_step_start .............................. 6
417
+ q_lora_rank ..................................... None
418
+ qk_head_dim ..................................... 128
419
+ qk_l2_norm ...................................... False
420
+ qk_layernorm .................................... False
421
+ qk_pos_emb_head_dim ............................. 64
422
+ query_in_block_prob ............................. 0.1
423
+ rampup_batch_size ............................... None
424
+ rank ............................................ 0
425
+ recompute_granularity ........................... None
426
+ recompute_method ................................ None
427
+ recompute_modules ............................... None
428
+ recompute_num_layers ............................ None
429
+ record_memory_history ........................... False
430
+ relative_attention_max_distance ................. 128
431
+ relative_attention_num_buckets .................. 32
432
+ replication ..................................... False
433
+ replication_factor .............................. 2
434
+ replication_jump ................................ None
435
+ rerun_mode ...................................... validate_results
436
+ reset_attention_mask ............................ False
437
+ reset_position_ids .............................. False
438
+ result_rejected_tracker_filename ................ None
439
+ retriever_report_topk_accuracies ................ []
440
+ retriever_score_scaling ......................... False
441
+ retriever_seq_length ............................ 256
442
+ retro_add_retriever ............................. False
443
+ retro_attention_gate ............................ 1
444
+ retro_cyclic_train_iters ........................ None
445
+ retro_encoder_attention_dropout ................. 0.1
446
+ retro_encoder_hidden_dropout .................... 0.1
447
+ retro_encoder_layers ............................ 2
448
+ retro_num_neighbors ............................. 2
449
+ retro_num_retrieved_chunks ...................... 2
450
+ retro_project_dir ............................... None
451
+ retro_verify_neighbor_count ..................... True
452
+ reuse_grad_buf_for_mxfp8_param_ag ............... False
453
+ rope_scaling_factor ............................. 8.0
454
+ rope_type ....................................... None
455
+ rotary_base ..................................... 10000
456
+ rotary_interleaved .............................. False
457
+ rotary_percent .................................. 1.0
458
+ rotary_scaling_factor ........................... 1.0
459
+ rotary_seq_len_interpolation_factor ............. None
460
+ run_workload_inspector_server ................... False
461
+ sample_rate ..................................... 1.0
462
+ save ............................................ None
463
+ save_interval ................................... None
464
+ save_retain_interval ............................ None
465
+ scatter_gather_tensors_in_pipeline .............. True
466
+ seed ............................................ 1234
467
+ seq_length ...................................... 4096
468
+ sequence_parallel ............................... False
469
+ sft ............................................. False
470
+ sft_tokenizer_prompt_format ..................... nemotron-h-aligned
471
+ sgd_momentum .................................... 0.9
472
+ short_seq_prob .................................. 0.1
473
+ skip_train ...................................... False
474
+ skipped_train_samples ........................... 0
475
+ spec ............................................ None
476
+ split ........................................... 100,0,0
477
+ squared_relu .................................... False
478
+ start_weight_decay .............................. 0.1
479
+ straggler_ctrlr_port ............................ 65535
480
+ straggler_minmax_count .......................... 1
481
+ suggested_communication_unit_size ............... None
482
+ swiglu .......................................... True
483
+ swin_backbone_type .............................. tiny
484
+ symmetric_ar_type ............................... None
485
+ te_rng_tracker .................................. True
486
+ tensor_model_parallel_size ...................... 1
487
+ tensorboard_dir ................................. None
488
+ tensorboard_log_interval ........................ 1
489
+ tensorboard_queue_size .......................... 1000
490
+ test_data_path .................................. None
491
+ test_mode ....................................... False
492
+ tiktoken_num_special_tokens ..................... 1000
493
+ tiktoken_pattern ................................ None
494
+ tiktoken_special_tokens ......................... None
495
+ timing_log_level ................................ 0
496
+ timing_log_option ............................... minmax
497
+ titles_data_path ................................ None
498
+ tokenizer_model ................................. None
499
+ tokenizer_type .................................. GPT2BPETokenizer
500
+ torch_fsdp2_reshard_after_forward ............... True
501
+ tp_comm_bootstrap_backend ....................... nccl
502
+ tp_comm_bulk_dgrad .............................. True
503
+ tp_comm_bulk_wgrad .............................. True
504
+ tp_comm_overlap ................................. False
505
+ tp_comm_overlap_ag .............................. True
506
+ tp_comm_overlap_cfg ............................. None
507
+ tp_comm_overlap_rs .............................. True
508
+ tp_comm_overlap_rs_dgrad ........................ False
509
+ tp_comm_split_ag ................................ True
510
+ tp_comm_split_rs ................................ True
511
+ train_data_path ................................. None
512
+ train_iters ..................................... 10
513
+ train_samples ................................... None
514
+ train_sync_interval ............................. None
515
+ transformer_impl ................................ transformer_engine
516
+ transformer_pipeline_model_parallel_size ........ 1
517
+ untie_embeddings_and_output_weights ............. True
518
+ use_checkpoint_args ............................. False
519
+ use_checkpoint_opt_param_scheduler .............. False
520
+ use_cpu_initialization .......................... None
521
+ use_custom_fsdp ................................. False
522
+ use_dist_ckpt ................................... True
523
+ use_dist_ckpt_deprecated ........................ False
524
+ use_distributed_optimizer ....................... True
525
+ use_flash_attn .................................. False
526
+ use_legacy_models ............................... False
527
+ use_mp_args_from_checkpoint_args ................ False
528
+ use_one_sent_docs ............................... False
529
+ use_persistent_ckpt_worker ...................... False
530
+ use_precision_aware_optimizer ................... False
531
+ use_pytorch_profiler ............................ False
532
+ use_ring_exchange_p2p ........................... False
533
+ use_rope_scaling ................................ False
534
+ use_rotary_position_embeddings .................. False
535
+ use_sharp ....................................... False
536
+ use_tokenizer_model_from_checkpoint_args ........ True
537
+ use_torch_fsdp2 ................................. False
538
+ use_torch_optimizer_for_cpu_offload ............. False
539
+ use_tp_pp_dp_mapping ............................ False
540
+ v_head_dim ...................................... 128
541
+ valid_data_path ................................. None
542
+ variable_seq_lengths ............................ False
543
+ virtual_pipeline_model_parallel_size ............ None
544
+ vision_backbone_type ............................ vit
545
+ vision_pretraining .............................. False
546
+ vision_pretraining_type ......................... classify
547
+ vocab_extra_ids ................................. 0
548
+ vocab_file ...................................... /workspace/data/wikitext/gpt2-vocab.json
549
+ vocab_size ...................................... None
550
+ wandb_exp_name ..................................
551
+ wandb_project ...................................
552
+ wandb_save_dir ..................................
553
+ weight_decay .................................... 0.1
554
+ weight_decay_incr_style ......................... constant
555
+ wgrad_deferral_limit ............................ 0
556
+ world_size ...................................... 8
557
+ yaml_cfg ........................................ None
558
+ -------------------- end of arguments ---------------------
559
+ INFO:megatron.core.num_microbatches_calculator:setting number of microbatches to constant 1
560
+ > building GPT2BPETokenizer tokenizer ...
561
+ > padded vocab (size: 50257) with 47 dummy tokens (new size: 50304)
562
+ WARNING:megatron.core.rerun_state_machine:RerunStateMachine initialized in mode RerunMode.VALIDATE_RESULTS
563
+ > initializing torch distributed ...
564
+ /workspace/Megatron-MoE-Echo/megatron/training/arguments.py:831: UserWarning: Disabling sequence parallelism because tensor model parallelism is disabled
565
+ warnings.warn("Disabling sequence parallelism because tensor model parallelism is disabled")
566
+ /workspace/Megatron-MoE-Echo/megatron/training/arguments.py:831: UserWarning: Disabling sequence parallelism because tensor model parallelism is disabled
567
+ warnings.warn("Disabling sequence parallelism because tensor model parallelism is disabled")
568
+ /workspace/Megatron-MoE-Echo/megatron/training/arguments.py:831: UserWarning: Disabling sequence parallelism because tensor model parallelism is disabled
569
+ warnings.warn("Disabling sequence parallelism because tensor model parallelism is disabled")
570
+ /workspace/Megatron-MoE-Echo/megatron/training/arguments.py:831: UserWarning: Disabling sequence parallelism because tensor model parallelism is disabled
571
+ warnings.warn("Disabling sequence parallelism because tensor model parallelism is disabled")
572
+ /workspace/Megatron-MoE-Echo/megatron/training/arguments.py:831: UserWarning: Disabling sequence parallelism because tensor model parallelism is disabled
573
+ warnings.warn("Disabling sequence parallelism because tensor model parallelism is disabled")
574
+ /workspace/Megatron-MoE-Echo/megatron/training/arguments.py:831: UserWarning: Disabling sequence parallelism because tensor model parallelism is disabled
575
+ warnings.warn("Disabling sequence parallelism because tensor model parallelism is disabled")
576
+ /workspace/Megatron-MoE-Echo/megatron/training/arguments.py:831: UserWarning: Disabling sequence parallelism because tensor model parallelism is disabled
577
+ warnings.warn("Disabling sequence parallelism because tensor model parallelism is disabled")
578
+ WARNING: one_logger package is required to enable e2e metrics tracking. please go to https://confluence.nvidia.com/display/MLWFO/Package+Repositories for details to install it
579
+ [Gloo] Rank 1 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
580
+ [Gloo] Rank 5 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
581
+ [Gloo] Rank 0 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
582
+ [Gloo] Rank 2 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
583
+ [Gloo] Rank 3 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
584
+ [Gloo] Rank 4 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
585
+ [Gloo] Rank 6 is connected to [Gloo] Rank 7 peer ranks. Expected number of connected peer ranks is : 77 is connected to
586
+ 7 peer ranks. Expected number of connected peer ranks is : 7
587
+ [Gloo] Rank 0 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
588
+ [Gloo] Rank [Gloo] Rank 1 is connected to 27 is connected to peer ranks. 7Expected number of connected peer ranks is : peer ranks. 7Expected number of connected peer ranks is : 7
589
+
590
+ [Gloo] Rank [Gloo] Rank 4 is connected to 7 peer ranks. 3Expected number of connected peer ranks is : is connected to 77 peer ranks.
591
+ Expected number of connected peer ranks is : 7
592
+ [Gloo] Rank 6 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
593
+ [Gloo] Rank 5 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
594
+ [Gloo] Rank 7 is connected to 7 peer ranks. Expected number of connected peer ranks is : 7
595
+ [Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
596
+ > initialized tensor model parallel with size 1
597
+ > initialized pipeline model parallel with size 1
598
+ > setting random seeds to 1234 ...
599
+ [Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
600
+ [Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
601
+ [Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
602
+ [Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
603
+ [Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
604
+ [Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
605
+ [Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
606
+ [rank2]:[W921 09:09:46.980057629 ProcessGroupNCCL.cpp:4789] [PG ID 0 PG GUID 0 Rank 2] using GPU 2 as device used by this process is currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. You can specify device_id in init_process_group() to force use of a particular device.
607
+ [rank1]:[W921 09:09:46.984640212 ProcessGroupNCCL.cpp:4789] [PG ID 0 PG GUID 0 Rank 1] using GPU 1 as device used by this process is currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. You can specify device_id in init_process_group() to force use of a particular device.
608
+ [rank7]:[W921 09:09:46.984718369 ProcessGroupNCCL.cpp:4789] [PG ID 0 PG GUID 0 Rank 7] using GPU 7 as device used by this process is currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. You can specify device_id in init_process_group() to force use of a particular device.
609
+ [rank6]:[W921 09:09:46.988524699 ProcessGroupNCCL.cpp:4789] [PG ID 0 PG GUID 0 Rank 6] using GPU 6 as device used by this process is currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. You can specify device_id in init_process_group() to force use of a particular device.
610
+ [rank5]:[W921 09:09:46.988562317 ProcessGroupNCCL.cpp:4789] [PG ID 0 PG GUID 0 Rank 5] using GPU 5 as device used by this process is currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. You can specify device_id in init_process_group() to force use of a particular device.
611
+ [rank3]:[W921 09:09:46.988579122 ProcessGroupNCCL.cpp:4789] [PG ID 0 PG GUID 0 Rank 3] using GPU 3 as device used by this process is currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. You can specify device_id in init_process_group() to force use of a particular device.
612
+ [rank4]:[W921 09:09:46.988726900 ProcessGroupNCCL.cpp:4789] [PG ID 0 PG GUID 0 Rank 4] using GPU 4 as device used by this process is currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. You can specify device_id in init_process_group() to force use of a particular device.
613
+ > compiling dataset index builder ...
614
+ make: Entering directory '/workspace/Megatron-MoE-Echo/megatron/core/datasets'
615
+ make: Nothing to be done for 'default'.
616
+ make: Leaving directory '/workspace/Megatron-MoE-Echo/megatron/core/datasets'
617
+ >>> done with dataset index builder. Compilation time: 0.430 seconds
618
+ WARNING: constraints for invoking optimized fused softmax kernel are not met. We default back to unfused kernel invocations.
619
+ > compiling and loading fused kernels ...
620
+ [rank0]:[W921 09:09:46.617521422 ProcessGroupNCCL.cpp:4789] [PG ID 0 PG GUID 0 Rank 0] using GPU 0 as device used by this process is currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. You can specify device_id in init_process_group() to force use of a particular device.
621
+ >>> done with compiling and loading fused kernels. Compilation time: 3.608 seconds
622
+ time to initialize megatron (seconds): 9.516
623
+ [after megatron is initialized] datetime: 2026-09-21 09:09:52
624
+ building GPT model ...
625
+ /usr/local/lib/python3.12/dist-packages/transformer_engine/pytorch/cpu_offload.py:708: DeprecationWarning: Offloading weights is deprecated. Using offload_weights=True does not have any effect.
626
+ warnings.warn(
627
+ /usr/local/lib/python3.12/dist-packages/transformer_engine/pytorch/cpu_offload.py:708: DeprecationWarning: Offloading weights is deprecated. Using offload_weights=True does not have any effect.
628
+ warnings.warn(
629
+ /usr/local/lib/python3.12/dist-packages/transformer_engine/pytorch/cpu_offload.py:708: DeprecationWarning: Offloading weights is deprecated. Using offload_weights=True does not have any effect.
630
+ warnings.warn(
631
+ /usr/local/lib/python3.12/dist-packages/transformer_engine/pytorch/cpu_offload.py:708: DeprecationWarning: Offloading weights is deprecated. Using offload_weights=True does not have any effect.
632
+ warnings.warn(
633
+ /usr/local/lib/python3.12/dist-packages/transformer_engine/pytorch/cpu_offload.py:708: DeprecationWarning: Offloading weights is deprecated. Using offload_weights=True does not have any effect.
634
+ warnings.warn(
635
+ /usr/local/lib/python3.12/dist-packages/transformer_engine/pytorch/cpu_offload.py:708: DeprecationWarning: Offloading weights is deprecated. Using offload_weights=True does not have any effect.
636
+ warnings.warn(
637
+ /usr/local/lib/python3.12/dist-packages/transformer_engine/pytorch/cpu_offload.py:708: DeprecationWarning: Offloading weights is deprecated. Using offload_weights=True does not have any effect.
638
+ warnings.warn(
639
+ /usr/local/lib/python3.12/dist-packages/transformer_engine/pytorch/cpu_offload.py:708: DeprecationWarning: Offloading weights is deprecated. Using offload_weights=True does not have any effect.
640
+ warnings.warn(
641
+ > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 1281951744
642
+ INFO:megatron.core.distributed.distributed_data_parallel:Setting up DistributedDataParallel with config DistributedDataParallelConfig(grad_reduce_in_fp32=True, overlap_grad_reduce=True, overlap_param_gather=True, align_param_gather=False, use_distributed_optimizer=True, num_distributed_optimizer_instances=1, check_for_nan_in_grad=False, check_for_large_grads=False, bucket_size=40000000, pad_buckets_for_high_nccl_busbw=False, average_in_collective=False, fp8_param_gather=False, reuse_grad_buf_for_mxfp8_param_ag=False, use_custom_fsdp=False, data_parallel_sharding_strategy='no_shard', gradient_reduce_div_fusion=True, suggested_communication_unit_size=None, preserve_fp32_weights=True, keep_fp8_transpose_cache_when_using_custom_fsdp=False, nccl_ub=False, fsdp_double_buffer=False)
643
+ INFO:megatron.core.distributed.param_and_grad_buffer:Number of buckets for gradient all-reduce / reduce-scatter: 7
644
+ Params for bucket 1 (103022592 elements, 103022592 padded size):
645
+ module.output_layer.weight
646
+ Params for bucket 2 (46542848 elements, 46542848 padded size):
647
+ module.decoder.final_layernorm.weight
648
+ module.decoder.layers.15.self_attention.linear_qkv.weight
649
+ module.decoder.layers.15.pre_mlp_layernorm.weight
650
+ module.decoder.layers.14.mlp.router.weight
651
+ module.decoder.layers.14.self_attention.linear_qkv.layer_norm_weight
652
+ module.decoder.layers.14.self_attention.linear_proj.weight
653
+ module.decoder.layers.13.pre_mlp_layernorm.weight
654
+ module.decoder.layers.13.self_attention.linear_qkv.weight
655
+ module.decoder.layers.15.mlp.router.weight
656
+ module.decoder.layers.15.self_attention.linear_qkv.layer_norm_weight
657
+ module.decoder.layers.15.self_attention.linear_proj.weight
658
+ module.decoder.layers.14.pre_mlp_layernorm.weight
659
+ module.decoder.layers.14.self_attention.linear_qkv.weight
660
+ module.decoder.layers.13.mlp.router.weight
661
+ Params for bucket 3 (50737152 elements, 50737152 padded size):
662
+ module.decoder.layers.12.mlp.router.weight
663
+ module.decoder.layers.12.self_attention.linear_qkv.layer_norm_weight
664
+ module.decoder.layers.12.self_attention.linear_proj.weight
665
+ module.decoder.layers.11.mlp.router.weight
666
+ module.decoder.layers.11.self_attention.linear_qkv.layer_norm_weight
667
+ module.decoder.layers.11.self_attention.linear_proj.weight
668
+ module.decoder.layers.10.pre_mlp_layernorm.weight
669
+ module.decoder.layers.10.self_attention.linear_qkv.weight
670
+ module.decoder.layers.13.self_attention.linear_qkv.layer_norm_weight
671
+ module.decoder.layers.13.self_attention.linear_proj.weight
672
+ module.decoder.layers.12.pre_mlp_layernorm.weight
673
+ module.decoder.layers.12.self_attention.linear_qkv.weight
674
+ module.decoder.layers.11.pre_mlp_layernorm.weight
675
+ module.decoder.layers.11.self_attention.linear_qkv.weight
676
+ module.decoder.layers.10.mlp.router.weight
677
+ Params for bucket 4 (50737152 elements, 50737152 padded size):
678
+ module.decoder.layers.9.mlp.router.weight
679
+ module.decoder.layers.9.self_attention.linear_qkv.layer_norm_weight
680
+ module.decoder.layers.9.self_attention.linear_proj.weight
681
+ module.decoder.layers.8.pre_mlp_layernorm.weight
682
+ module.decoder.layers.8.self_attention.linear_qkv.weight
683
+ module.decoder.layers.7.mlp.router.weight
684
+ module.decoder.layers.10.self_attention.linear_qkv.layer_norm_weight
685
+ module.decoder.layers.10.self_attention.linear_proj.weight
686
+ module.decoder.layers.9.pre_mlp_layernorm.weight
687
+ module.decoder.layers.9.self_attention.linear_qkv.weight
688
+ module.decoder.layers.8.mlp.router.weight
689
+ module.decoder.layers.8.self_attention.linear_qkv.layer_norm_weight
690
+ module.decoder.layers.8.self_attention.linear_proj.weight
691
+ module.decoder.layers.7.pre_mlp_layernorm.weight
692
+ module.decoder.layers.7.self_attention.linear_qkv.weight
693
+ Params for bucket 5 (50737152 elements, 50737152 padded size):
694
+ module.decoder.layers.7.self_attention.linear_qkv.layer_norm_weight
695
+ module.decoder.layers.6.mlp.router.weight
696
+ module.decoder.layers.7.self_attention.linear_proj.weight
697
+ module.decoder.layers.6.self_attention.linear_qkv.layer_norm_weight
698
+ module.decoder.layers.6.self_attention.linear_proj.weight
699
+ module.decoder.layers.5.pre_mlp_layernorm.weight
700
+ module.decoder.layers.5.self_attention.linear_qkv.weight
701
+ module.decoder.layers.5.self_attention.linear_proj.weight
702
+ module.decoder.layers.4.mlp.router.weight
703
+ module.decoder.layers.6.self_attention.linear_qkv.weight
704
+ module.decoder.layers.6.pre_mlp_layernorm.weight
705
+ module.decoder.layers.5.mlp.router.weight
706
+ module.decoder.layers.5.self_attention.linear_qkv.layer_norm_weight
707
+ module.decoder.layers.4.pre_mlp_layernorm.weight
708
+ module.decoder.layers.4.self_attention.linear_qkv.weight
709
+ Params for bucket 6 (50737152 elements, 50737152 padded size):
710
+ module.decoder.layers.4.self_attention.linear_qkv.layer_norm_weight
711
+ module.decoder.layers.3.self_attention.linear_qkv.weight
712
+ module.decoder.layers.3.pre_mlp_layernorm.weight
713
+ module.decoder.layers.4.self_attention.linear_proj.weight
714
+ module.decoder.layers.2.self_attention.linear_qkv.layer_norm_weight
715
+ module.decoder.layers.2.self_attention.linear_proj.weight
716
+ module.decoder.layers.1.pre_mlp_layernorm.weight
717
+ module.decoder.layers.1.self_attention.linear_qkv.weight
718
+ module.decoder.layers.3.mlp.router.weight
719
+ module.decoder.layers.3.self_attention.linear_qkv.layer_norm_weight
720
+ module.decoder.layers.3.self_attention.linear_proj.weight
721
+ module.decoder.layers.2.mlp.router.weight
722
+ module.decoder.layers.2.pre_mlp_layernorm.weight
723
+ module.decoder.layers.2.self_attention.linear_qkv.weight
724
+ module.decoder.layers.1.mlp.router.weight
725
+ Params for bucket 7 (124131328 elements, 124131328 padded size):
726
+ module.decoder.layers.1.self_attention.linear_qkv.layer_norm_weight
727
+ module.decoder.layers.0.self_attention.linear_qkv.weight
728
+ module.decoder.layers.0.pre_mlp_layernorm.weight
729
+ module.decoder.layers.1.self_attention.linear_proj.weight
730
+ module.decoder.layers.0.self_attention.linear_qkv.layer_norm_weight
731
+ module.decoder.layers.0.mlp.router.weight
732
+ module.decoder.layers.0.self_attention.linear_proj.weight
733
+ module.embedding.word_embeddings.weight
734
+ INFO:megatron.core.distributed.param_and_grad_buffer:Number of buckets for gradient all-reduce / reduce-scatter: 20
735
+ Params for bucket 1 (41943040 elements, 41943040 padded size):
736
+ module.decoder.layers.15.mlp.experts.linear_fc2.weight3
737
+ module.decoder.layers.15.mlp.experts.linear_fc2.weight1
738
+ module.decoder.layers.15.mlp.experts.linear_fc2.weight5
739
+ module.decoder.layers.15.mlp.experts.linear_fc2.weight7
740
+ module.decoder.layers.15.mlp.experts.linear_fc1.weight0
741
+ module.decoder.layers.15.mlp.experts.linear_fc1.weight2
742
+ module.decoder.layers.15.mlp.experts.linear_fc1.weight4
743
+ module.decoder.layers.15.mlp.experts.linear_fc2.weight0
744
+ module.decoder.layers.15.mlp.experts.linear_fc2.weight2
745
+ module.decoder.layers.15.mlp.experts.linear_fc2.weight4
746
+ module.decoder.layers.15.mlp.experts.linear_fc2.weight6
747
+ module.decoder.layers.15.mlp.experts.linear_fc1.weight1
748
+ module.decoder.layers.15.mlp.experts.linear_fc1.weight3
749
+ module.decoder.layers.15.mlp.experts.linear_fc1.weight5
750
+ Params for bucket 2 (41943040 elements, 41943040 padded size):
751
+ module.decoder.layers.15.mlp.experts.linear_fc1.weight6
752
+ module.decoder.layers.14.mlp.experts.linear_fc2.weight0
753
+ module.decoder.layers.14.mlp.experts.linear_fc2.weight2
754
+ module.decoder.layers.14.mlp.experts.linear_fc2.weight4
755
+ module.decoder.layers.14.mlp.experts.linear_fc2.weight6
756
+ module.decoder.layers.14.mlp.experts.linear_fc1.weight1
757
+ module.decoder.layers.14.mlp.experts.linear_fc1.weight3
758
+ module.decoder.layers.14.mlp.experts.linear_fc2.weight1
759
+ module.decoder.layers.15.mlp.experts.linear_fc1.weight7
760
+ module.decoder.layers.14.mlp.experts.linear_fc2.weight3
761
+ module.decoder.layers.14.mlp.experts.linear_fc2.weight5
762
+ module.decoder.layers.14.mlp.experts.linear_fc2.weight7
763
+ module.decoder.layers.14.mlp.experts.linear_fc1.weight0
764
+ module.decoder.layers.14.mlp.experts.linear_fc1.weight2
765
+ Params for bucket 3 (41943040 elements, 41943040 padded size):
766
+ module.decoder.layers.14.mlp.experts.linear_fc1.weight7
767
+ module.decoder.layers.14.mlp.experts.linear_fc1.weight5
768
+ module.decoder.layers.13.mlp.experts.linear_fc2.weight1
769
+ module.decoder.layers.13.mlp.experts.linear_fc2.weight3
770
+ module.decoder.layers.13.mlp.experts.linear_fc2.weight5
771
+ module.decoder.layers.13.mlp.experts.linear_fc2.weight7
772
+ module.decoder.layers.13.mlp.experts.linear_fc1.weight0
773
+ module.decoder.layers.14.mlp.experts.linear_fc1.weight4
774
+ module.decoder.layers.14.mlp.experts.linear_fc1.weight6
775
+ module.decoder.layers.13.mlp.experts.linear_fc2.weight0
776
+ module.decoder.layers.13.mlp.experts.linear_fc2.weight2
777
+ module.decoder.layers.13.mlp.experts.linear_fc2.weight4
778
+ module.decoder.layers.13.mlp.experts.linear_fc2.weight6
779
+ module.decoder.layers.13.mlp.experts.linear_fc1.weight1
780
+ Params for bucket 4 (41943040 elements, 41943040 padded size):
781
+ module.decoder.layers.13.mlp.experts.linear_fc1.weight2
782
+ module.decoder.layers.13.mlp.experts.linear_fc1.weight6
783
+ module.decoder.layers.13.mlp.experts.linear_fc1.weight4
784
+ module.decoder.layers.12.mlp.experts.linear_fc2.weight0
785
+ module.decoder.layers.12.mlp.experts.linear_fc2.weight2
786
+ module.decoder.layers.12.mlp.experts.linear_fc2.weight4
787
+ module.decoder.layers.12.mlp.experts.linear_fc2.weight6
788
+ module.decoder.layers.13.mlp.experts.linear_fc1.weight3
789
+ module.decoder.layers.13.mlp.experts.linear_fc1.weight5
790
+ module.decoder.layers.13.mlp.experts.linear_fc1.weight7
791
+ module.decoder.layers.12.mlp.experts.linear_fc2.weight1
792
+ module.decoder.layers.12.mlp.experts.linear_fc2.weight3
793
+ module.decoder.layers.12.mlp.experts.linear_fc2.weight5
794
+ module.decoder.layers.12.mlp.experts.linear_fc2.weight7
795
+ Params for bucket 5 (41943040 elements, 41943040 padded size):
796
+ module.decoder.layers.12.mlp.experts.linear_fc1.weight3
797
+ module.decoder.layers.12.mlp.experts.linear_fc1.weight1
798
+ module.decoder.layers.12.mlp.experts.linear_fc1.weight5
799
+ module.decoder.layers.12.mlp.experts.linear_fc1.weight7
800
+ module.decoder.layers.11.mlp.experts.linear_fc2.weight1
801
+ module.decoder.layers.11.mlp.experts.linear_fc2.weight3
802
+ module.decoder.layers.12.mlp.experts.linear_fc1.weight0
803
+ module.decoder.layers.12.mlp.experts.linear_fc1.weight4
804
+ module.decoder.layers.12.mlp.experts.linear_fc1.weight2
805
+ module.decoder.layers.12.mlp.experts.linear_fc1.weight6
806
+ module.decoder.layers.11.mlp.experts.linear_fc2.weight0
807
+ module.decoder.layers.11.mlp.experts.linear_fc2.weight2
808
+ Params for bucket 6 (41943040 elements, 41943040 padded size):
809
+ module.decoder.layers.11.mlp.experts.linear_fc2.weight7
810
+ module.decoder.layers.11.mlp.experts.linear_fc2.weight5
811
+ module.decoder.layers.11.mlp.experts.linear_fc1.weight1
812
+ module.decoder.layers.11.mlp.experts.linear_fc1.weight3
813
+ module.decoder.layers.11.mlp.experts.linear_fc1.weight5
814
+ module.decoder.layers.11.mlp.experts.linear_fc1.weight7
815
+ module.decoder.layers.11.mlp.experts.linear_fc2.weight4
816
+ module.decoder.layers.11.mlp.experts.linear_fc2.weight6
817
+ module.decoder.layers.11.mlp.experts.linear_fc1.weight0
818
+ module.decoder.layers.11.mlp.experts.linear_fc1.weight2
819
+ module.decoder.layers.11.mlp.experts.linear_fc1.weight4
820
+ module.decoder.layers.11.mlp.experts.linear_fc1.weight6
821
+ Params for bucket 7 (41943040 elements, 41943040 padded size):
822
+ module.decoder.layers.10.mlp.experts.linear_fc2.weight0
823
+ module.decoder.layers.10.mlp.experts.linear_fc2.weight2
824
+ module.decoder.layers.10.mlp.experts.linear_fc2.weight4
825
+ module.decoder.layers.10.mlp.experts.linear_fc2.weight6
826
+ module.decoder.layers.10.mlp.experts.linear_fc1.weight1
827
+ module.decoder.layers.10.mlp.experts.linear_fc1.weight3
828
+ module.decoder.layers.10.mlp.experts.linear_fc1.weight5
829
+ module.decoder.layers.10.mlp.experts.linear_fc2.weight3
830
+ module.decoder.layers.10.mlp.experts.linear_fc2.weight1
831
+ module.decoder.layers.10.mlp.experts.linear_fc2.weight5
832
+ module.decoder.layers.10.mlp.experts.linear_fc2.weight7
833
+ module.decoder.layers.10.mlp.experts.linear_fc1.weight0
834
+ module.decoder.layers.10.mlp.experts.linear_fc1.weight2
835
+ module.decoder.layers.10.mlp.experts.linear_fc1.weight4
836
+ Params for bucket 8 (41943040 elements, 41943040 padded size):
837
+ module.decoder.layers.9.mlp.experts.linear_fc2.weight0
838
+ module.decoder.layers.9.mlp.experts.linear_fc2.weight2
839
+ module.decoder.layers.10.mlp.experts.linear_fc1.weight7
840
+ module.decoder.layers.9.mlp.experts.linear_fc2.weight4
841
+ module.decoder.layers.9.mlp.experts.linear_fc2.weight6
842
+ module.decoder.layers.9.mlp.experts.linear_fc1.weight1
843
+ module.decoder.layers.9.mlp.experts.linear_fc1.weight3
844
+ module.decoder.layers.10.mlp.experts.linear_fc1.weight6
845
+ module.decoder.layers.9.mlp.experts.linear_fc2.weight1
846
+ module.decoder.layers.9.mlp.experts.linear_fc2.weight3
847
+ module.decoder.layers.9.mlp.experts.linear_fc2.weight5
848
+ module.decoder.layers.9.mlp.experts.linear_fc2.weight7
849
+ module.decoder.layers.9.mlp.experts.linear_fc1.weight0
850
+ module.decoder.layers.9.mlp.experts.linear_fc1.weight2
851
+ Params for bucket 9 (41943040 elements, 41943040 padded size):
852
+ module.decoder.layers.9.mlp.experts.linear_fc1.weight7
853
+ module.decoder.layers.9.mlp.experts.linear_fc1.weight5
854
+ module.decoder.layers.8.mlp.experts.linear_fc2.weight1
855
+ module.decoder.layers.8.mlp.experts.linear_fc2.weight3
856
+ module.decoder.layers.8.mlp.experts.linear_fc2.weight5
857
+ module.decoder.layers.8.mlp.experts.linear_fc2.weight7
858
+ module.decoder.layers.8.mlp.experts.linear_fc1.weight0
859
+ module.decoder.layers.9.mlp.experts.linear_fc1.weight4
860
+ module.decoder.layers.8.mlp.experts.linear_fc2.weight0
861
+ module.decoder.layers.9.mlp.experts.linear_fc1.weight6
862
+ module.decoder.layers.8.mlp.experts.linear_fc2.weight2
863
+ module.decoder.layers.8.mlp.experts.linear_fc2.weight4
864
+ module.decoder.layers.8.mlp.experts.linear_fc2.weight6
865
+ module.decoder.layers.8.mlp.experts.linear_fc1.weight1
866
+ Params for bucket 10 (41943040 elements, 41943040 padded size):
867
+ module.decoder.layers.8.mlp.experts.linear_fc1.weight2
868
+ module.decoder.layers.8.mlp.experts.linear_fc1.weight4
869
+ module.decoder.layers.8.mlp.experts.linear_fc1.weight6
870
+ module.decoder.layers.7.mlp.experts.linear_fc2.weight0
871
+ module.decoder.layers.7.mlp.experts.linear_fc2.weight2
872
+ module.decoder.layers.7.mlp.experts.linear_fc2.weight4
873
+ module.decoder.layers.7.mlp.experts.linear_fc2.weight6
874
+ module.decoder.layers.8.mlp.experts.linear_fc1.weight5
875
+ module.decoder.layers.8.mlp.experts.linear_fc1.weight3
876
+ module.decoder.layers.8.mlp.experts.linear_fc1.weight7
877
+ module.decoder.layers.7.mlp.experts.linear_fc2.weight1
878
+ module.decoder.layers.7.mlp.experts.linear_fc2.weight3
879
+ module.decoder.layers.7.mlp.experts.linear_fc2.weight5
880
+ module.decoder.layers.7.mlp.experts.linear_fc2.weight7
881
+ Params for bucket 11 (41943040 elements, 41943040 padded size):
882
+ module.decoder.layers.7.mlp.experts.linear_fc1.weight1
883
+ module.decoder.layers.7.mlp.experts.linear_fc1.weight3
884
+ module.decoder.layers.7.mlp.experts.linear_fc1.weight5
885
+ module.decoder.layers.7.mlp.experts.linear_fc1.weight7
886
+ module.decoder.layers.6.mlp.experts.linear_fc2.weight1
887
+ module.decoder.layers.6.mlp.experts.linear_fc2.weight3
888
+ module.decoder.layers.7.mlp.experts.linear_fc1.weight0
889
+ module.decoder.layers.7.mlp.experts.linear_fc1.weight4
890
+ module.decoder.layers.7.mlp.experts.linear_fc1.weight2
891
+ module.decoder.layers.7.mlp.experts.linear_fc1.weight6
892
+ module.decoder.layers.6.mlp.experts.linear_fc2.weight0
893
+ module.decoder.layers.6.mlp.experts.linear_fc2.weight2
894
+ Params for bucket 12 (41943040 elements, 41943040 padded size):
895
+ module.decoder.layers.6.mlp.experts.linear_fc2.weight5
896
+ module.decoder.layers.6.mlp.experts.linear_fc1.weight0
897
+ module.decoder.layers.6.mlp.experts.linear_fc2.weight7
898
+ module.decoder.layers.6.mlp.experts.linear_fc1.weight2
899
+ module.decoder.layers.6.mlp.experts.linear_fc1.weight4
900
+ module.decoder.layers.6.mlp.experts.linear_fc1.weight6
901
+ module.decoder.layers.6.mlp.experts.linear_fc2.weight4
902
+ module.decoder.layers.6.mlp.experts.linear_fc2.weight6
903
+ module.decoder.layers.6.mlp.experts.linear_fc1.weight1
904
+ module.decoder.layers.6.mlp.experts.linear_fc1.weight3
905
+ module.decoder.layers.6.mlp.experts.linear_fc1.weight5
906
+ module.decoder.layers.6.mlp.experts.linear_fc1.weight7
907
+ Params for bucket 13 (41943040 elements, 41943040 padded size):
908
+ module.decoder.layers.5.mlp.experts.linear_fc2.weight1
909
+ module.decoder.layers.5.mlp.experts.linear_fc2.weight3
910
+ module.decoder.layers.5.mlp.experts.linear_fc2.weight5
911
+ module.decoder.layers.5.mlp.experts.linear_fc2.weight7
912
+ module.decoder.layers.5.mlp.experts.linear_fc1.weight1
913
+ module.decoder.layers.5.mlp.experts.linear_fc1.weight3
914
+ module.decoder.layers.5.mlp.experts.linear_fc1.weight5
915
+ module.decoder.layers.5.mlp.experts.linear_fc2.weight0
916
+ module.decoder.layers.5.mlp.experts.linear_fc2.weight4
917
+ module.decoder.layers.5.mlp.experts.linear_fc2.weight2
918
+ module.decoder.layers.5.mlp.experts.linear_fc2.weight6
919
+ module.decoder.layers.5.mlp.experts.linear_fc1.weight0
920
+ module.decoder.layers.5.mlp.experts.linear_fc1.weight2
921
+ module.decoder.layers.5.mlp.experts.linear_fc1.weight4
922
+ Params for bucket 14 (41943040 elements, 41943040 padded size):
923
+ module.decoder.layers.4.mlp.experts.linear_fc2.weight2
924
+ module.decoder.layers.5.mlp.experts.linear_fc1.weight7
925
+ module.decoder.layers.4.mlp.experts.linear_fc2.weight0
926
+ module.decoder.layers.4.mlp.experts.linear_fc2.weight4
927
+ module.decoder.layers.4.mlp.experts.linear_fc2.weight6
928
+ module.decoder.layers.4.mlp.experts.linear_fc1.weight1
929
+ module.decoder.layers.4.mlp.experts.linear_fc1.weight3
930
+ module.decoder.layers.5.mlp.experts.linear_fc1.weight6
931
+ module.decoder.layers.4.mlp.experts.linear_fc2.weight1
932
+ module.decoder.layers.4.mlp.experts.linear_fc2.weight3
933
+ module.decoder.layers.4.mlp.experts.linear_fc2.weight5
934
+ module.decoder.layers.4.mlp.experts.linear_fc2.weight7
935
+ module.decoder.layers.4.mlp.experts.linear_fc1.weight0
936
+ module.decoder.layers.4.mlp.experts.linear_fc1.weight2
937
+ Params for bucket 15 (41943040 elements, 41943040 padded size):
938
+ module.decoder.layers.4.mlp.experts.linear_fc1.weight7
939
+ module.decoder.layers.4.mlp.experts.linear_fc1.weight5
940
+ module.decoder.layers.3.mlp.experts.linear_fc2.weight1
941
+ module.decoder.layers.3.mlp.experts.linear_fc2.weight3
942
+ module.decoder.layers.3.mlp.experts.linear_fc2.weight5
943
+ module.decoder.layers.3.mlp.experts.linear_fc2.weight7
944
+ module.decoder.layers.3.mlp.experts.linear_fc1.weight0
945
+ module.decoder.layers.4.mlp.experts.linear_fc1.weight4
946
+ module.decoder.layers.3.mlp.experts.linear_fc2.weight0
947
+ module.decoder.layers.4.mlp.experts.linear_fc1.weight6
948
+ module.decoder.layers.3.mlp.experts.linear_fc2.weight2
949
+ module.decoder.layers.3.mlp.experts.linear_fc2.weight4
950
+ module.decoder.layers.3.mlp.experts.linear_fc2.weight6
951
+ module.decoder.layers.3.mlp.experts.linear_fc1.weight1
952
+ Params for bucket 16 (41943040 elements, 41943040 padded size):
953
+ module.decoder.layers.3.mlp.experts.linear_fc1.weight2
954
+ module.decoder.layers.3.mlp.experts.linear_fc1.weight6
955
+ module.decoder.layers.3.mlp.experts.linear_fc1.weight4
956
+ module.decoder.layers.2.mlp.experts.linear_fc2.weight0
957
+ module.decoder.layers.2.mlp.experts.linear_fc2.weight2
958
+ module.decoder.layers.2.mlp.experts.linear_fc2.weight4
959
+ module.decoder.layers.2.mlp.experts.linear_fc2.weight6
960
+ module.decoder.layers.3.mlp.experts.linear_fc1.weight5
961
+ module.decoder.layers.3.mlp.experts.linear_fc1.weight3
962
+ module.decoder.layers.3.mlp.experts.linear_fc1.weight7
963
+ module.decoder.layers.2.mlp.experts.linear_fc2.weight1
964
+ module.decoder.layers.2.mlp.experts.linear_fc2.weight3
965
+ module.decoder.layers.2.mlp.experts.linear_fc2.weight5
966
+ module.decoder.layers.2.mlp.experts.linear_fc2.weight7
967
+ Params for bucket 17 (41943040 elements, 41943040 padded size):
968
+ module.decoder.layers.2.mlp.experts.linear_fc1.weight0
969
+ module.decoder.layers.2.mlp.experts.linear_fc1.weight4
970
+ module.decoder.layers.2.mlp.experts.linear_fc1.weight2
971
+ module.decoder.layers.2.mlp.experts.linear_fc1.weight6
972
+ module.decoder.layers.1.mlp.experts.linear_fc2.weight0
973
+ module.decoder.layers.1.mlp.experts.linear_fc2.weight1
974
+ module.decoder.layers.1.mlp.experts.linear_fc2.weight2
975
+ module.decoder.layers.2.mlp.experts.linear_fc1.weight3
976
+ module.decoder.layers.2.mlp.experts.linear_fc1.weight1
977
+ module.decoder.layers.2.mlp.experts.linear_fc1.weight5
978
+ module.decoder.layers.2.mlp.experts.linear_fc1.weight7
979
+ module.decoder.layers.1.mlp.experts.linear_fc2.weight3
980
+ Params for bucket 18 (41943040 elements, 41943040 padded size):
981
+ module.decoder.layers.1.mlp.experts.linear_fc1.weight0
982
+ module.decoder.layers.1.mlp.experts.linear_fc2.weight7
983
+ module.decoder.layers.1.mlp.experts.linear_fc1.weight1
984
+ module.decoder.layers.1.mlp.experts.linear_fc1.weight4
985
+ module.decoder.layers.1.mlp.experts.linear_fc1.weight6
986
+ module.decoder.layers.1.mlp.experts.linear_fc2.weight4
987
+ module.decoder.layers.1.mlp.experts.linear_fc2.weight6
988
+ module.decoder.layers.1.mlp.experts.linear_fc2.weight5
989
+ module.decoder.layers.1.mlp.experts.linear_fc1.weight2
990
+ module.decoder.layers.1.mlp.experts.linear_fc1.weight3
991
+ module.decoder.layers.1.mlp.experts.linear_fc1.weight5
992
+ module.decoder.layers.1.mlp.experts.linear_fc1.weight7
993
+ Params for bucket 19 (41943040 elements, 41943040 padded size):
994
+ module.decoder.layers.0.mlp.experts.linear_fc2.weight3
995
+ module.decoder.layers.0.mlp.experts.linear_fc2.weight1
996
+ module.decoder.layers.0.mlp.experts.linear_fc2.weight5
997
+ module.decoder.layers.0.mlp.experts.linear_fc2.weight7
998
+ module.decoder.layers.0.mlp.experts.linear_fc1.weight1
999
+ module.decoder.layers.0.mlp.experts.linear_fc1.weight2
1000
+ module.decoder.layers.0.mlp.experts.linear_fc2.weight0
1001
+ module.decoder.layers.0.mlp.experts.linear_fc2.weight4
1002
+ module.decoder.layers.0.mlp.experts.linear_fc2.weight2
1003
+ module.decoder.layers.0.mlp.experts.linear_fc2.weight6
1004
+ module.decoder.layers.0.mlp.experts.linear_fc1.weight0
1005
+ module.decoder.layers.0.mlp.experts.linear_fc1.weight3
1006
+ module.decoder.layers.0.mlp.experts.linear_fc1.weight4
1007
+ module.decoder.layers.0.mlp.experts.linear_fc1.weight5
1008
+ Params for bucket 20 (8388608 elements, 8388608 padded size):
1009
+ module.decoder.layers.0.mlp.experts.linear_fc1.weight6
1010
+ module.decoder.layers.0.mlp.experts.linear_fc1.weight7
1011
+ INFO:megatron.core.optimizer:Setting up optimizer with config OptimizerConfig(optimizer='adam', lr=0.0001, min_lr=1e-05, decoupled_lr=None, decoupled_min_lr=None, weight_decay=0.1, fp8_recipe='mxfp8', fp16=False, bf16=True, reuse_grad_buf_for_mxfp8_param_ag=False, params_dtype=torch.bfloat16, use_precision_aware_optimizer=False, store_param_remainders=True, main_grads_dtype=torch.float32, main_params_dtype=torch.float32, exp_avg_dtype=torch.float32, exp_avg_sq_dtype=torch.float32, loss_scale=None, initial_loss_scale=4294967296, min_loss_scale=1.0, loss_scale_window=1000, hysteresis=2, adam_beta1=0.9, adam_beta2=0.999, adam_eps=1e-08, sgd_momentum=0.9, use_distributed_optimizer=True, overlap_param_gather_with_optimizer_step=False, optimizer_cpu_offload=False, optimizer_offload_fraction=1.0, use_torch_optimizer_for_cpu_offload=False, overlap_cpu_optimizer_d2h_h2d=False, pin_cpu_grads=True, pin_cpu_params=True, clip_grad=1.0, log_num_zeros_in_grad=False, barrier_with_L1_time=True, timers=<megatron.core.timers.Timers object at 0x729be6e88f20>, config_logger_dir='')
1012
+ INFO:megatron.core.optimizer_param_scheduler:> learning rate decay style: cosine
1013
+ [after model, optimizer, and learning rate scheduler are built] datetime: 2026-09-21 09:09:55
1014
+ > building train, validation, and test datasets ...
1015
+ > datasets target sizes (minimum size):
1016
+ train: 80
1017
+ validation: 0
1018
+ test: 0
1019
+ INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split_matrix = [(0, 1.0), None, None]
1020
+ > building train, validation, and test datasets for GPT ...
1021
+ INFO:megatron.core.datasets.blended_megatron_dataset_builder:Building GPTDataset splits with sizes=(80, 0, 0) and config=GPTDatasetConfig(random_seed=1234, sequence_length=4096, blend=(['/workspace/data/wikitext/wikitext103_gpt2_text_document'], None), blend_per_split=None, split='100,0,0', split_matrix=[(0, 1.0), None, None], num_dataset_builder_threads=1, path_to_cache=None, mmap_bin_files=True, mock=False, tokenizer=<megatron.training.tokenizer.tokenizer._GPT2BPETokenizer object at 0x729be6ffacc0>, mid_level_dataset_surplus=0.005, reset_position_ids=False, reset_attention_mask=False, eod_mask_loss=False, create_attention_mask=True, drop_last_partial_validation_sequence=True, add_extra_token_to_sequence=True, object_storage_cache_path=None)
1022
+ INFO:megatron.core.datasets.indexed_dataset:Load the _IndexReader from /workspace/data/wikitext/wikitext103_gpt2_text_document.idx
1023
+ INFO:megatron.core.datasets.indexed_dataset: Extract the sequence lengths
1024
+ INFO:megatron.core.datasets.indexed_dataset: Extract the sequence pointers
1025
+ INFO:megatron.core.datasets.indexed_dataset: Extract the document indices
1026
+ INFO:megatron.core.datasets.indexed_dataset:> total number of sequences: 1165029
1027
+ INFO:megatron.core.datasets.indexed_dataset:> total number of documents: 1801350
1028
+ INFO:megatron.core.datasets.gpt_dataset:Load the GPTDataset train indices
1029
+ INFO:megatron.core.datasets.gpt_dataset: Load the document index from 71e99bc7c46bc3412b339a272e0f9cf3-GPTDataset-train-document_index.npy
1030
+ INFO:megatron.core.datasets.gpt_dataset: Load the sample index from 71e99bc7c46bc3412b339a272e0f9cf3-GPTDataset-train-sample_index.npy
1031
+ INFO:megatron.core.datasets.gpt_dataset: Load the shuffle index from 71e99bc7c46bc3412b339a272e0f9cf3-GPTDataset-train-shuffle_index.npy
1032
+ INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 29073
1033
+ > finished creating GPT datasets ...
1034
+ [after dataloaders are built] datetime: 2026-09-21 09:09:55
1035
+ done with setup ...
1036
+ (min, max) time across ranks (ms):
1037
+ model-and-optimizer-setup ......................: (2782.23, 2783.22)
1038
+ train/valid/test-data-iterators-setup ..........: (89.52, 110.27)
1039
+ training ...
1040
+ Overwriting rerun_state_machine.current_iteration from -1 to 0...
1041
+ [before the start of training step] datetime: 2026-09-21 09:09:55
1042
+ INFO:moonep.api:MoonEP comm buffer sizes:
1043
+ config : R=8, S=4096, H=2048, E=64, K=8, NvS=34800, NvS_padded=34816
1044
+ sm_count : num_sms=32, num_sms_dedup=148
1045
+ hidden_buffer_size : 136.00 MiB
1046
+ meta_buffer_size : 2.00 MiB
1047
+ local_temp_buffer_size : 1.26 MiB
1048
+ total_size : 139.26 MiB
1049
+ [MoonEP selfcheck 1/4] dropped groups empty OK | covered rows [0, 1, 2, 3, 4, 5, 6, 7, 64, 65, 66, 67, 68, 69, 70, 71] sizes [16512, 5632, 1152, 2944, 1920, 3840, 1024, 384, 128, 0, 0, 0, 0, 0, 0, 0] | duplicated experts [[44, -1, -1, -1, -1, -1, -1, -1], [18, -1, -1, -1, -1, -1, -1, -1], [52, -1, -1, -1, -1, -1, -1, -1], [44, -1, -1, -1, -1, -1, -1, -1], [27, -1, -1, -1, -1, -1, -1, -1], [-1, -1, -1, -1, -1, -1, -1, -1], [27, -1, -1, -1, -1, -1, -1, -1], [19, -1, -1, -1, -1, -1, -1, -1]]
1050
+ [MoonEP selfcheck 2/4] dropped groups empty OK | covered rows [0, 1, 2, 3, 4, 5, 6, 7, 64, 65, 66, 67, 68, 69, 70, 71] sizes [256, 4736, 8960, 256, 256, 768, 8704, 384, 9088, 0, 0, 0, 0, 0, 0, 0] | duplicated experts [[27, -1, -1, -1, -1, -1, -1, -1], [29, -1, -1, -1, -1, -1, -1, -1], [31, -1, -1, -1, -1, -1, -1, -1], [39, -1, -1, -1, -1, -1, -1, -1], [-1, -1, -1, -1, -1, -1, -1, -1], [28, -1, -1, -1, -1, -1, -1, -1], [31, -1, -1, -1, -1, -1, -1, -1], [27, -1, -1, -1, -1, -1, -1, -1]]
1051
+ [MoonEP selfcheck 3/4] dropped groups empty OK | covered rows [0, 1, 2, 3, 4, 5, 6, 7, 64, 65, 66, 67, 68, 69, 70, 71] sizes [128, 384, 128, 1280, 128, 128, 1408, 384, 27904, 1664, 0, 0, 0, 0, 0, 0] | duplicated experts [[9, 8, -1, -1, -1, -1, -1, -1], [61, -1, -1, -1, -1, -1, -1, -1], [60, -1, -1, -1, -1, -1, -1, -1], [-1, -1, -1, -1, -1, -1, -1, -1], [27, -1, -1, -1, -1, -1, -1, -1], [27, -1, -1, -1, -1, -1, -1, -1], [60, -1, -1, -1, -1, -1, -1, -1], [27, -1, -1, -1, -1, -1, -1, -1]]
1052
+ [MoonEP selfcheck 4/4] dropped groups empty OK | covered rows [0, 1, 2, 3, 4, 5, 6, 7, 64, 65, 66, 67, 68, 69, 70, 71] sizes [128, 7296, 4352, 1408, 8320, 4864, 128, 6528, 256, 0, 0, 0, 0, 0, 0, 0] | duplicated experts [[9, -1, -1, -1, -1, -1, -1, -1], [-1, -1, -1, -1, -1, -1, -1, -1], [42, -1, -1, -1, -1, -1, -1, -1], [52, -1, -1, -1, -1, -1, -1, -1], [1, -1, -1, -1, -1, -1, -1, -1], [34, -1, -1, -1, -1, -1, -1, -1], [44, -1, -1, -1, -1, -1, -1, -1], [46, -1, -1, -1, -1, -1, -1, -1]]
1053
+ Number of parameters in transformer block in billions: 6.71
1054
+ [2026-09-21 09:10:15] iteration 1/ 10 | consumed samples: 8 | elapsed time per iteration (ms): 19274.7 | throughput per GPU (TFLOP/s/GPU): 1.7 | learning rate: 2.000000E-05 | global batch size: 8 | lm loss: 1.118505E+01 | loss scale: 1.0 | grad norm: 35.489 | number of skipped iterations: 0 | number of nan iterations: 0 |Number of parameters in embedding layers in billions: 0.21
1055
+
1056
+ Total number of parameters in billions: 6.92
1057
+ Number of parameters in most loaded shard in billions: 6.9171
1058
+ Theoretical memory footprints: weight and optimizer=49474.72 MB
1059
+ [Rank 0] (after 1 iterations) memory (MB) | allocated: 16643.6826171875 | max allocated: 16999.8642578125 | reserved: 17484.0 | max reserved: 17484.0
1060
+ [2026-09-21 09:10:15] iteration 2/ 10 | consumed samples: 16 | elapsed time per iteration (ms): 238.1 | throughput per GPU (TFLOP/s/GPU): 135.3 | learning rate: 4.000000E-05 | global batch size: 8 | lm loss: 1.120947E+01 | loss scale: 1.0 | grad norm: 36.302 | number of skipped iterations: 0 | number of nan iterations: 0 |
1061
+ [2026-09-21 09:10:15] iteration 3/ 10 | consumed samples: 24 | elapsed time per iteration (ms): 192.6 | throughput per GPU (TFLOP/s/GPU): 167.3 | learning rate: 6.000000E-05 | global batch size: 8 | lm loss: 1.006579E+01 | loss scale: 1.0 | grad norm: 101.390 | number of skipped iterations: 0 | number of nan iterations: 0 |
1062
+ Capture CUDA graph for training!!!
1063
+ CUDA graph capture done!!!
1064
+ [2026-09-21 09:10:16] iteration 4/ 10 | consumed samples: 32 | elapsed time per iteration (ms): 703.8 | throughput per GPU (TFLOP/s/GPU): 45.8 | learning rate: 8.000000E-05 | global batch size: 8 | lm loss: 1.023741E+01 | loss scale: 1.0 | grad norm: 28.774 | number of skipped iterations: 0 | number of nan iterations: 0 |
1065
+ [2026-09-21 09:10:16] iteration 5/ 10 | consumed samples: 40 | elapsed time per iteration (ms): 107.6 | throughput per GPU (TFLOP/s/GPU): 299.4 | learning rate: 1.000000E-04 | global batch size: 8 | lm loss: 1.073561E+01 | loss scale: 1.0 | grad norm: 20.803 | number of skipped iterations: 0 | number of nan iterations: 0 |
1066
+ [2026-09-21 09:10:16] iteration 6/ 10 | consumed samples: 48 | elapsed time per iteration (ms): 107.0 | throughput per GPU (TFLOP/s/GPU): 301.1 | learning rate: 7.750000E-05 | global batch size: 8 | lm loss: 9.625093E+00 | loss scale: 1.0 | grad norm: 8.176 | number of skipped iterations: 0 | number of nan iterations: 0 |
1067
+ [2026-09-21 09:10:19] iteration 7/ 10 | consumed samples: 56 | elapsed time per iteration (ms): 2878.0 | throughput per GPU (TFLOP/s/GPU): 11.2 | learning rate: 3.250000E-05 | global batch size: 8 | lm loss: 9.077366E+00 | loss scale: 1.0 | grad norm: 16.538 | number of skipped iterations: 0 | number of nan iterations: 0 |
1068
+ [2026-09-21 09:10:19] iteration 8/ 10 | consumed samples: 64 | elapsed time per iteration (ms): 123.3 | throughput per GPU (TFLOP/s/GPU): 261.4 | learning rate: 1.000000E-05 | global batch size: 8 | lm loss: 8.817106E+00 | loss scale: 1.0 | grad norm: 8.752 | number of skipped iterations: 0 | number of nan iterations: 0 |
1069
+ Capture range started in the application.
1070
+ Capture range ended in the application.
1071
+ Generating '/tmp/nsys-report-d000.qdstrm'
1072
+
1073
+
1074
+ [2026-09-21 09:11:00] iteration 9/ 10 | consumed samples: 72 | elapsed time per iteration (ms): 40781.4 | throughput per GPU (TFLOP/s/GPU): 0.8 | learning rate: 1.000000E-05 | global batch size: 8 | lm loss: 8.709936E+00 | loss scale: 1.0 | grad norm: 3.776 | number of skipped iterations: 0 | number of nan iterations: 0 |
1075
+ [2026-09-21 09:11:00] iteration 10/ 10 | consumed samples: 80 | elapsed time per iteration (ms): 117.0 | throughput per GPU (TFLOP/s/GPU): 275.5 | learning rate: 1.000000E-05 | global batch size: 8 | lm loss: 8.653146E+00 | loss scale: 1.0 | grad norm: 3.500 | number of skipped iterations: 0 | number of nan iterations: 0 |
1076
+ [after training is done] datetime: 2026-09-21 09:11:00
1077
+ [rank0]:[W921 09:11:01.827459267 ProcessGroupNCCL.cpp:1505] 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())
1078
+ [rank3]:[W921 09:11:02.320133821 ProcessGroupNCCL.cpp:1505] 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())
1079
+ Collection stop is not allowed in this state.
1080
+ [rank4]:[W921 09:11:02.711637598 ProcessGroupNCCL.cpp:1505] 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())
1081
+ [rank2]:[W921 09:11:03.998989338 ProcessGroupNCCL.cpp:1505] 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())
1082
+ [rank6]:[W921 09:11:03.271879905 ProcessGroupNCCL.cpp:1505] 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())
1083
+ [rank5]:[W921 09:11:03.275148389 ProcessGroupNCCL.cpp:1505] 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())
1084
+ [rank7]:[W921 09:11:03.455467791 ProcessGroupNCCL.cpp:1505] 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())
1085
+ [rank1]:[W921 09:11:04.982932473 ProcessGroupNCCL.cpp:1505] 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())
1086
+ [rank3]:[W921 09:11:06.596467870 observer.cpp:116] Warning: Leaked callback handle: 1 (function operator())
1087
+ pure virtual method called
1088
+ terminate called without an active exception
1089
+ [rank6]:[W921 09:11:06.745868023 observer.cpp:116] Warning: Leaked callback handle: 1 (function operator())
1090
+ pure virtual method called
1091
+ terminate called without an active exception
1092
+ [rank0]:[W921 09:11:06.776881054 observer.cpp:116] Warning: Leaked callback handle: 1 (function operator())
1093
+ [rank4]:[W921 09:11:06.782840931 observer.cpp:116] Warning: Leaked callback handle: 1 (function operator())
1094
+ pure virtual method called
1095
+ terminate called without an active exception
1096
+ pure virtual method called
1097
+ terminate called without an active exception
1098
+ W0921 09:11:07.252000 806907 torch/distributed/elastic/multiprocessing/api.py:900] Sending process 807662 closing signal SIGTERM
1099
+ W0921 09:11:07.253000 806907 torch/distributed/elastic/multiprocessing/api.py:900] Sending process 807663 closing signal SIGTERM
1100
+ W0921 09:11:07.253000 806907 torch/distributed/elastic/multiprocessing/api.py:900] Sending process 807664 closing signal SIGTERM
1101
+ W0921 09:11:07.254000 806907 torch/distributed/elastic/multiprocessing/api.py:900] Sending process 807666 closing signal SIGTERM
1102
+ W0921 09:11:07.255000 806907 torch/distributed/elastic/multiprocessing/api.py:900] Sending process 807667 closing signal SIGTERM
1103
+ W0921 09:11:07.255000 806907 torch/distributed/elastic/multiprocessing/api.py:900] Sending process 807668 closing signal SIGTERM
1104
+ W0921 09:11:07.256000 806907 torch/distributed/elastic/multiprocessing/api.py:900] Sending process 807669 closing signal SIGTERM
1105
+ E0921 09:11:08.587000 806907 torch/distributed/elastic/multiprocessing/api.py:874] failed (exitcode: -6) local_rank: 3 (pid: 807665) of binary: /usr/bin/python
1106
+ Traceback (most recent call last):
1107
+ File "/usr/local/bin/torchrun", line 8, in <module>
1108
+ sys.exit(main())
1109
+ ^^^^^^
1110
+ File "/usr/local/lib/python3.12/dist-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 355, in wrapper
1111
+ return f(*args, **kwargs)
1112
+ ^^^^^^^^^^^^^^^^^^
1113
+ File "/usr/local/lib/python3.12/dist-packages/torch/distributed/run.py", line 892, in main
1114
+ run(args)
1115
+ File "/usr/local/lib/python3.12/dist-packages/torch/distributed/run.py", line 883, in run
1116
+ elastic_launch(
1117
+ File "/usr/local/lib/python3.12/dist-packages/torch/distributed/launcher/api.py", line 139, in __call__
1118
+ return launch_agent(self._config, self._entrypoint, list(args))
1119
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1120
+ File "/usr/local/lib/python3.12/dist-packages/torch/distributed/launcher/api.py", line 270, in launch_agent
1121
+ raise ChildFailedError(
1122
+ torch.distributed.elastic.multiprocessing.errors.ChildFailedError:
1123
+ ============================================================
1124
+ pretrain_gpt.py FAILED
1125
+ ------------------------------------------------------------
1126
+ Failures:
1127
+ <NO_OTHER_FAILURES>
1128
+ ------------------------------------------------------------
1129
+ Root Cause (first observed failure):
1130
+ [0]:
1131
+ time : 2026-09-21_09:11:07
1132
+ host : sp-aiinfra-interactive-aghosh-8-mb99k-8xb200-moe-cg-0
1133
+ rank : 3 (local_rank: 3)
1134
+ exitcode : -6 (pid: 807665)
1135
+ error_file: <N/A>
1136
+ traceback : Signal 6 (SIGABRT) received by PID 807665
1137
+ ============================================================
1138
+ Generated:
1139
+ /workspace/granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.nsys-rep
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/verify_trace_ranks.log ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ report : /workspace/granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.nsys-rep
2
+ kernels : 73596
3
+ devices : 8 of 8 -> [0, 1, 2, 3, 4, 5, 6, 7]
4
+ processes: 1
5
+ markers : 8704 moonep_trace__ kernels on devices [0, 1, 2, 3, 4, 5, 6, 7]
6
+ PASS: all ranks present and backend markers emitted
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/results.tsv ADDED
@@ -0,0 +1 @@
 
 
1
+ config backend policy slots seq gbs cap recompute rep status iter_ms_mean iter_ms_min peak_mib run_dir
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/trace_extension_build.log ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Could not find an up-to-date installation of `packaging`. License expressions might not be validated. To enforce validation, please install `packaging>=24.2`.
2
+ /usr/local/lib/python3.12/dist-packages/setuptools/config/_apply_pyprojecttoml.py:82: SetuptoolsDeprecationWarning: `project.license` as a TOML table is deprecated
3
+ !!
4
+
5
+ ********************************************************************************
6
+ Please use a simple string containing a SPDX expression for `project.license`. You can also use `project.license-files`. (Both options available on setuptools>=77.0.0).
7
+
8
+ This deprecation is overdue, please update your project and remove deprecated
9
+ calls to avoid build errors in the future.
10
+
11
+ See https://packaging.python.org/en/latest/guides/writing-pyproject-toml/#license for details.
12
+ ********************************************************************************
13
+
14
+ !!
15
+ corresp(dist, value, root_dir)
16
+ /usr/local/lib/python3.12/dist-packages/setuptools/config/_apply_pyprojecttoml.py:61: SetuptoolsDeprecationWarning: License classifiers are deprecated.
17
+ !!
18
+
19
+ ********************************************************************************
20
+ Please consider removing the following classifiers in favor of a SPDX license expression:
21
+
22
+ License :: OSI Approved :: BSD License
23
+
24
+ See https://packaging.python.org/en/latest/guides/writing-pyproject-toml/#license for details.
25
+ ********************************************************************************
26
+
27
+ !!
28
+ dist._finalize_license_expression()
29
+ /usr/local/lib/python3.12/dist-packages/setuptools/dist.py:759: SetuptoolsDeprecationWarning: License classifiers are deprecated.
30
+ !!
31
+
32
+ ********************************************************************************
33
+ Please consider removing the following classifiers in favor of a SPDX license expression:
34
+
35
+ License :: OSI Approved :: BSD License
36
+
37
+ See https://packaging.python.org/en/latest/guides/writing-pyproject-toml/#license for details.
38
+ ********************************************************************************
39
+
40
+ !!
41
+ self._finalize_license_expression()
42
+ running build_ext
43
+ building 'megatron.core.datasets.helpers_cpp' extension
44
+ Emitting ninja build file /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/build.ninja...
45
+ Compiling objects...
46
+ Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N)
47
+ ninja: no work to do.
48
+ x86_64-linux-gnu-g++ -fno-strict-overflow -Wsign-compare -DNDEBUG -g -O2 -Wall -shared -Wl,-O1 -Wl,-Bsymbolic-functions /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/megatron/core/datasets/helpers.o -L/usr/lib/x86_64-linux-gnu -o build/lib.linux-x86_64-cpython-312/megatron/core/datasets/helpers_cpp.cpython-312-x86_64-linux-gnu.so
49
+ building 'megatron.core.transformer.moe._moonep_trace_C' extension
50
+ Emitting ninja build file /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/build.ninja...
51
+ Compiling objects...
52
+ Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N)
53
+ [1/2] c++ -MMD -MF /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/megatron/core/transformer/moe/moonep_trace_ops.o.d -fno-strict-overflow -Wsign-compare -DNDEBUG -g -O2 -Wall -fPIC -I/usr/local/lib/python3.12/dist-packages/torch/include -I/usr/local/lib/python3.12/dist-packages/torch/include/torch/csrc/api/include -I/usr/local/cuda/include -I/usr/include/python3.12 -c -c /workspace/Megatron-MoE-Echo/megatron/core/transformer/moe/moonep_trace_ops.cpp -o /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/megatron/core/transformer/moe/moonep_trace_ops.o -O3 -std=c++17 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=_moonep_trace_C
54
+ [2/2] /usr/local/cuda/bin/nvcc --generate-dependencies-with-compile --dependency-output /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/megatron/core/transformer/moe/moonep_trace_kernels.o.d -I/usr/local/lib/python3.12/dist-packages/torch/include -I/usr/local/lib/python3.12/dist-packages/torch/include/torch/csrc/api/include -I/usr/local/cuda/include -I/usr/include/python3.12 -c -c /workspace/Megatron-MoE-Echo/megatron/core/transformer/moe/moonep_trace_kernels.cu -o /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/megatron/core/transformer/moe/moonep_trace_kernels.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -O3 -std=c++17 -lineinfo -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=_moonep_trace_C -gencode=arch=compute_100,code=sm_100 -gencode=arch=compute_120,code=compute_120 -gencode=arch=compute_120,code=sm_120 -gencode=arch=compute_75,code=sm_75 -gencode=arch=compute_80,code=sm_80 -gencode=arch=compute_86,code=sm_86 -gencode=arch=compute_90,code=sm_90
55
+ x86_64-linux-gnu-g++ -fno-strict-overflow -Wsign-compare -DNDEBUG -g -O2 -Wall -shared -Wl,-O1 -Wl,-Bsymbolic-functions /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/megatron/core/transformer/moe/moonep_trace_kernels.o /workspace/Megatron-MoE-Echo/build/temp.linux-x86_64-cpython-312/megatron/core/transformer/moe/moonep_trace_ops.o -L/usr/local/lib/python3.12/dist-packages/torch/lib -L/usr/local/cuda/lib64 -L/usr/lib/x86_64-linux-gnu -lc10 -ltorch -ltorch_cpu -ltorch_python -lcudart -lc10_cuda -ltorch_cuda -o build/lib.linux-x86_64-cpython-312/megatron/core/transformer/moe/_moonep_trace_C.cpython-312-x86_64-linux-gnu.so
56
+ copying build/lib.linux-x86_64-cpython-312/megatron/core/datasets/helpers_cpp.cpython-312-x86_64-linux-gnu.so -> megatron/core/datasets
57
+ copying build/lib.linux-x86_64-cpython-312/megatron/core/transformer/moe/_moonep_trace_C.cpython-312-x86_64-linux-gnu.so -> megatron/core/transformer/moe
58
+ load check passed: /workspace/Megatron-MoE-Echo/megatron/core/transformer/moe/_moonep_trace_C.cpython-312-x86_64-linux-gnu.so -> torch.ops.megatron_moonep_trace.fwd_moonep_forward_begin
granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/trace_results.tsv ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ model backend capacity status report async_overlap_json
2
+ OLMOE_1B_7B_L16 moonep none OK /workspace/granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async.nsys-rep /workspace/granularity_next_b200/tierb_models/traces/OLMOE_1B_7B_L16/moonep/tierbtrace/tierbtrace_gbs8/OLMOE_1B_7B_L16__moonep__matched__seq4096__gbs8__capnone__async/nsys/moonep_async_overlap.json