-
Notifications
You must be signed in to change notification settings - Fork 146
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add back 24.05 response sending path to fix performance #381
Merged
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
krishung5
force-pushed
the
krish-fix-perf
branch
from
October 4, 2024 00:23
df8f2db
to
1282598
Compare
Tabrizian
reviewed
Oct 4, 2024
Tabrizian
reviewed
Oct 4, 2024
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks Kris 🙏
Tabrizian
approved these changes
Oct 4, 2024
Is there anything holding up this PR? Looks like it should be ready to merge. We need this to intercept the 0.14 release of TRT-LLM. Code freeze just started today. |
kthui
approved these changes
Oct 8, 2024
krishung5
added a commit
that referenced
this pull request
Oct 9, 2024
* Add back 24.05 response sender path * Improve perf * Fix cleanup * Review comments * Fix up * Fix up * Fix response factory cleanup * Fix segfault * Fix error handling * Remove extra logs * Fix up, add comments * Address comment * Fix up --------- Co-authored-by: Iman Tabrizian <itabrizian@nvidia.com>
pvijayakrish
pushed a commit
that referenced
this pull request
Oct 15, 2024
* Add back 24.05 response sender path * Improve perf * Fix cleanup * Review comments * Fix up * Fix up * Fix response factory cleanup * Fix segfault * Fix error handling * Remove extra logs * Fix up, add comments * Address comment * Fix up --------- Co-authored-by: Iman Tabrizian <itabrizian@nvidia.com>
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.
(Continuing from #378)
This PR restores the response sending path where the response batch is used to send responses from the stub to the backend, which was the method used in 24.05. In this PR, the method was changed to use
response_sender.Send()
to send responses back to the backend fromexecute()
, with the goal of consolidating the two methods. However, we are observing performance regressions due to an extra IPC message in theresponse_sender.Send()
path. Since the stub must send theresponse_batch
to the backend regardless, the response returned fromexecute()
can simply be appended to theresponse_batch
, avoiding the extra IPC round.Results:
Before
After
24.05