Record comms input and output tensor information #1014
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
This pull request was exported from Phabricator. Differential Revision: D65785010 |
sraikund16
approved these changes
Nov 14, 2024
sanrise
added a commit
to sanrise/kineto
that referenced
this pull request
Nov 14, 2024
Summary: Just copy the output from NCCL metadata about tensor information. Allows for easier analysis of kernel memory access patterns in downstream tools. Reviewed By: sraikund16 Differential Revision: D65785010
sanrise
force-pushed
the
export-D65785010
branch
from
7d78560 to
40b9db2
Compare
|
This pull request was exported from Phabricator. Differential Revision: D65785010 |
sanrise
added a commit
to sanrise/kineto
that referenced
this pull request
Nov 14, 2024
Summary: Just copy the output from NCCL metadata about tensor information. Allows for easier analysis of kernel memory access patterns in downstream tools. Reviewed By: sraikund16 Differential Revision: D65785010
sanrise
force-pushed
the
export-D65785010
branch
from
40b9db2 to
1539c9a
Compare
|
This pull request was exported from Phabricator. Differential Revision: D65785010 |
sanrise
added a commit
to sanrise/kineto
that referenced
this pull request
Nov 15, 2024
Summary: Just copy the output from NCCL metadata about tensor information. Allows for easier analysis of kernel memory access patterns in downstream tools. Reviewed By: sraikund16 Differential Revision: D65785010
sanrise
force-pushed
the
export-D65785010
branch
from
1539c9a to
5d83c43
Compare
|
This pull request was exported from Phabricator. Differential Revision: D65785010 |
sanrise
added a commit
to sanrise/kineto
that referenced
this pull request
Nov 15, 2024
Summary: Just copy the output from NCCL metadata about tensor information. Allows for easier analysis of kernel memory access patterns in downstream tools. Reviewed By: sraikund16 Differential Revision: D65785010
sanrise
force-pushed
the
export-D65785010
branch
from
5d83c43 to
fa2d485
Compare
|
This pull request was exported from Phabricator. Differential Revision: D65785010 |
Summary: Just copy the output from NCCL metadata about tensor information. Allows for easier analysis of kernel memory access patterns in downstream tools. Reviewed By: sraikund16 Differential Revision: D65785010
sanrise
force-pushed
the
export-D65785010
branch
from
fa2d485 to
50ae474
Compare
|
This pull request was exported from Phabricator. Differential Revision: D65785010 |
|
This pull request has been merged in d9673d2. |
|
This pull request has been reverted by 158d409. |
sanrise
added a commit
to sanrise/kineto
that referenced
this pull request
Nov 25, 2024
Summary: Reverts a rollback D66458621 Revert the Kineto rollback, this would have partially solved the issue since this part controls transmission of the metadata to the corresponding kernel. but record_param_comms in pytorch is the real issue and was still recording this metadata and would still make an invalid trace JSON when working with GPUs>30 (our truncation case). Differential Revision: D66475394
facebook-github-bot
pushed a commit
that referenced
this pull request
Nov 26, 2024
Summary: Pull Request resolved: #1017 Reverts a rollback D66458621 Revert the Kineto rollback, this would have partially solved the issue since this part controls transmission of the metadata to the corresponding kernel. but record_param_comms in pytorch is the real issue and was still recording this metadata and would still make an invalid trace JSON when working with GPUs>30 (our truncation case). Reviewed By: sraikund16 Differential Revision: D66475394 fbshipit-source-id: 5b781ccb27fa898a1a6496c72733f72fd31c822e
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary: Just copy the output from NCCL metadata about tensor information. Allows for easier analysis of kernel memory access patterns in downstream tools.
Differential Revision: D65785010