perf(megatron): skip redundant weight load - #3397
Merged
Merged
Conversation
|
Auto-sync is disabled for ready for review pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
tdene
force-pushed
the
tde/skip_inference_weight_load
branch
from
July 29, 2026 06:12
bc65313 to
6507e43
Compare
tdene
added
the
CI:Lfast
Contributor
Author
|
/ok to test 6507e43 |
shanmugamr1992
previously approved these changes
Jul 29, 2026
tdene
force-pushed
the
tde/skip_inference_weight_load
branch
from
August 4, 2026 05:33
6507e43 to
aa96ed3
Compare
Contributor
Author
|
/ok to test 5b80bd4 |
cspades
previously approved these changes
Aug 4, 2026
shanmugamr1992
previously approved these changes
Aug 4, 2026
tdene
added
CI:L1
Contributor
Author
|
/ok to test 95d5a81 |
tdene
force-pushed
the
tde/skip_inference_weight_load
branch
from
August 4, 2026 22:59
95d5a81 to
e19d73c
Compare
Contributor
Author
|
/ok to test e19d73c |
Closes NVIDIA-NeMo#2615 Signed-off-by: Teodor-Dumitru Ene <teodord.ene@gmail.com>
Signed-off-by: Teodor-Dumitru Ene <teodord.ene@gmail.com>
Signed-off-by: Teodor-Dumitru Ene <teodord.ene@gmail.com>
tdene
force-pushed
the
tde/skip_inference_weight_load
branch
from
August 5, 2026 03:23
e19d73c to
91ccce7
Compare
Contributor
Author
|
/ok to test 91ccce7 |
terrykong
approved these changes
Aug 5, 2026
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do ?
#2355 updated Megatron Inference to be a full peer of vllm inference. Model conversion still happens twice, once on training workers, once on inference workers. It should happen only once, and just be streamed across.
Issues
List issues that this PR closes (syntax):
#2615
Usage
# Add a code snippet demonstrating how to use thisBefore your PR is "Ready for review"
Pre checks:
Additional Information