[dev] pull main 260220 by FDecaYed · Pull Request #3574 · NVIDIA/Megatron-LM

FDecaYed · 2026-02-25T01:51:48Z

What does this PR do ?

⚠️ For major changes (either in lines of code or in its impact), please make sure to first share a design doc with the team. If you're unsure what's the best way to do so, contact the @mcore-oncall.

Contribution process

flowchart LR
    A[Pre-checks] --> B[PR Tests]
    subgraph Code Review/Approval
        C1[Expert Review] --> C2[Final Review]
    end
    B --> C1
    C2 --> D[Merge]

Pre-checks

I want this PR in a versioned release and have added the appropriate Milestone (e.g., Core 0.8)
I have added relevant unit tests
I have added relevant functional tests
I have added proper typing to my code Typing guidelines
I have added relevant documentation
I have run the autoformatter.sh on my PR

Code review

The following process is enforced via the CODEOWNERS file for changes into megatron/core. For changes outside of megatron/core, it is up to the PR author whether or not to tag the Final Reviewer team.

For MRs into `main` branch

Feel free to message or comment the @mcore-oncall to help accelerate your merge into main. The less complex your PR is, the faster it will be approved and merged!

(Step 1): Add PR label `Expert Review`

(Step 2): Collect the expert reviewers reviews

Attach the Expert Review label when your PR is ready for review.
GitHub auto-assigns expert reviewers based on your changes. They will get notified and pick up your PR soon.

⚠️ Only proceed to the next step once all reviewers have approved, merge-conflict are resolved and the CI is passing.
Final Review might get declined if these requirements are not fulfilled.

(Step 3): Final Review

Add Final Review label
GitHub auto-assigns final reviewers based on your changes. They will get notified and pick up your PR soon.

(Optional Step 4): Cherry-pick into release branch

If this PR also needs to be merged into core_r* release branches, after this PR has been merged, select Cherry-pick to open a new PR into the release branch.

For MRs into `dev` branch

The proposed review process for `dev` branch is under active discussion.

MRs are mergable after one approval by either eharper@nvidia.com or zijiey@nvidia.com.

Merging your PR

Any member of core-adlr and core-nemo will be able to merge your PR.

Co-authored-by: Philip Petrakian <ppetrakian@nvidia.com> Co-authored-by: oliver könig <okoenig@nvidia.com>

Signed-off-by: oliver könig <okoenig@nvidia.com>

Signed-off-by: Charlie Truong <chtruong@nvidia.com>

Signed-off-by: oliver könig <okoenig@nvidia.com>

Co-authored-by: root <root@gpu-h100-0348.cm.cluster> Co-authored-by: root <root@gpu-h100-0193.cm.cluster> Co-authored-by: root <root@gpu-h100-0082.cm.cluster> Co-authored-by: root <root@gpu-h100-0495.cm.cluster> Co-authored-by: William Dykas <wdykas@cw-pdx-cs-001-vscode-02.cm.cluster> Co-authored-by: root <root@gpu-h100-0213.cm.cluster> Co-authored-by: root <root@gpu-h100-0435.cm.cluster> Co-authored-by: root <root@gpu-h100-0188.cm.cluster> Co-authored-by: root <root@gpu-h100-0032.cm.cluster> Co-authored-by: root <root@gpu-h100-0023.cm.cluster> Co-authored-by: root <root@gpu-h100-0368.cm.cluster> Co-authored-by: root <root@gpu-h100-0203.cm.cluster> Co-authored-by: root <root@gpu-h100-0229.cm.cluster> Co-authored-by: root <root@gpu-h100-0123.cm.cluster> Co-authored-by: root <root@gpu-h100-0217.cm.cluster> Co-authored-by: root <root@gpu-h100-0496.cm.cluster> Co-authored-by: root <root@gpu-h100-0261.cm.cluster>

…ernorm. (NVIDIA#2434) Co-authored-by: Yuzhong Wang <yuzhongw@nvidia.com>

Signed-off-by: Hongbin Liu <hongbinl@nvidia.com>

Signed-off-by: Youngeun Kwon <youngeunk@nvidia.com> Co-authored-by: Youngeun Kwon <youngeunk@nvidia.com>

Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>

This reverts commit ffbc43f.

Signed-off-by: oliver könig <okoenig@nvidia.com>

Signed-off-by: Jimmy Zhang <jiemingz@nvidia.com>

Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com> Co-authored-by: Xin Yao <xiny@nvidia.com>

Co-authored-by: Rabeeh Mahabadi <rkarimimahab@nb-hel-cs-001-vscode-02.cm.cluster> Co-authored-by: Sanjeev Satheesh <sasatheesh@nvidia.com> Co-authored-by: Deepak Narayanan <dnarayanan@nvidia.com>

Signed-off-by: Santosh Bhavani <santosh.bhavani@live.com> Co-authored-by: Xin Yao <xiny@nvidia.com>

…offload` (NVIDIA#2874)

Co-authored-by: Maanu Grover <109391026+maanug-nv@users.noreply.github.com>

Signed-off-by: Deepak Narayanan <dnarayanan@nvidia.com>

Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com>

…VIDIA#3194)" This reverts commit e836e62.

This reverts commit 300d1b6.

Co-authored-by: Antoni-Joan Solergibert <asolergibert@nvidia.com>

Signed-off-by: Hollow Man <hollowman@opensuse.org> Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com> Co-authored-by: Xin Yao <xiny@nvidia.com>

Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com>

…#3419)

Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>

…VIDIA#3424)

Signed-off-by: mikail <mkhona@nvidia.com>

copy-pr-bot · 2026-02-25T01:51:53Z

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

FDecaYed · 2026-02-25T01:52:20Z

/ok to test f70db20

FDecaYed · 2026-02-25T01:55:51Z

/ok to test 1ffec4e

FDecaYed · 2026-02-25T06:08:50Z

/ok to test dcc87fe

FDecaYed · 2026-02-25T07:22:48Z

/ok to test 6e1d003

FDecaYed · 2026-02-25T10:17:37Z

/ok to test 5b7a481

FDecaYed · 2026-02-25T15:36:03Z

/ok to test 6531920

FDecaYed · 2026-02-26T03:32:56Z

/ok to test 0ab47fa

chtruong814 · 2026-02-26T05:16:26Z

/ok to test 0ab47fa

HaochenYuan and others added 30 commits January 30, 2026 09:54

Fix for PR-2142 (NVIDIA#3165)

4cd9563

Co-authored-by: Philip Petrakian <ppetrakian@nvidia.com> Co-authored-by: oliver könig <okoenig@nvidia.com>

ci: Onboard more GB200 tests (NVIDIA#3145)

6de6362

Signed-off-by: oliver könig <okoenig@nvidia.com>

ci(hotfix): Alert for GB200 (NVIDIA#3168)

de15117

Signed-off-by: oliver könig <okoenig@nvidia.com>

Fix SFTDataset truncation bug (NVIDIA#3158)

7952d7e

Vitalyk/multiturn v2 (NVIDIA#3167)

b9ee19e

ci: Disable the api check for now (NVIDIA#3157)

b168849

Signed-off-by: Charlie Truong <chtruong@nvidia.com>

ci: Add DSv3 proxy (NVIDIA#3169)

a205538

Signed-off-by: oliver könig <okoenig@nvidia.com>

Update copy-pr-bot.yaml [skip ci]

fdc04f6

[Community][Main] fix(moe): Fix theoretical memory calculation of lay…

9ad5906

…ernorm. (NVIDIA#2434) Co-authored-by: Yuzhong Wang <yuzhongw@nvidia.com>

fix: Set --refit-method default to gloo (NVIDIA#3172)

5415e1d

[fix] Bug fix for offloading in evaluate() (NVIDIA#3043)

a976754

Signed-off-by: Hongbin Liu <hongbinl@nvidia.com>

Update copy-pr-bot.yaml [skip ci]

991c38f

cp: Fix: nccl-ub in ddp path (3181) into main (NVIDIA#3182)

5d0a7fd

Signed-off-by: Youngeun Kwon <youngeunk@nvidia.com> Co-authored-by: Youngeun Kwon <youngeunk@nvidia.com>

Miscellaneous inference cleanup (NVIDIA#2955)

ffbc43f

Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>

Revert "Miscellaneous inference cleanup (NVIDIA#2955)"

0fe3232

This reverts commit ffbc43f.

ci: Fix DSv3 (NVIDIA#3188)

69a5c63

Signed-off-by: oliver könig <okoenig@nvidia.com>

Fix missing argument in MoELayer.forward() (NVIDIA#3133)

2fadde8

Signed-off-by: Jimmy Zhang <jiemingz@nvidia.com>

Fix H2D stream synchronization in optimizer offload (NVIDIA#3140)

ae67076

Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com> Co-authored-by: Xin Yao <xiny@nvidia.com>

Add MTP support for hybrid models (NVIDIA#2363)

300d1b6

Co-authored-by: Rabeeh Mahabadi <rkarimimahab@nb-hel-cs-001-vscode-02.cm.cluster> Co-authored-by: Sanjeev Satheesh <sasatheesh@nvidia.com> Co-authored-by: Deepak Narayanan <dnarayanan@nvidia.com>

docs: improve Megatron-LM and Megatron Core descriptions (NVIDIA#3115)

dceb1fb

Signed-off-by: Santosh Bhavani <santosh.bhavani@live.com> Co-authored-by: Xin Yao <xiny@nvidia.com>

Handle step key correctly in checkpoint save with `--optimizer-cpu-…

f4502eb

…offload` (NVIDIA#2874)

mRoPE for MTP (NVIDIA#3114)

70719cd

Co-authored-by: Maanu Grover <109391026+maanug-nv@users.noreply.github.com>

Fix two minor bugs in MTP implementation for hybrid models (NVIDIA#3194)

e836e62

Signed-off-by: Deepak Narayanan <dnarayanan@nvidia.com>

Update README.md (NVIDIA#2111)

1362e4a

Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com>

Revert "Fix two minor bugs in MTP implementation for hybrid models (N…

31d0c87

…VIDIA#3194)" This reverts commit e836e62.

Revert "Add MTP support for hybrid models (NVIDIA#2363)"

a0cc8ca

This reverts commit 300d1b6.

Fix bug in SFTDataset (NVIDIA#3185)

50546da

Co-authored-by: Antoni-Joan Solergibert <asolergibert@nvidia.com>

Fix several syntax error (NVIDIA#3004)

dff4189

Signed-off-by: Hollow Man <hollowman@opensuse.org> Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com> Co-authored-by: Xin Yao <xiny@nvidia.com>

Fix for RL Test (NVIDIA#3148)

c4bea0a

Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com>

nschank and others added 5 commits February 24, 2026 21:30

Use copy_signature to preserve typing of pass-through methods (NVIDIA…

782e54b

…#3419)

Ensure type-checker understands use of Submodules in MTP (NVIDIA#3308)

3597312

Add mxfp8 quantization for inference linear layers (NVIDIA#3447)

44e27d0

Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>

Add single-process checkpoint save to avoid forked multiprocessing (N…

2cac78b

…VIDIA#3424)

Fixed fp32 residuals (NVIDIA#3504)

08857d9

Signed-off-by: mikail <mkhona@nvidia.com>

FDecaYed requested review from a team as code owners February 25, 2026 01:51

svcnvidia-nemo-ci added this to the Core 0.16 milestone Feb 25, 2026

copy-pr-bot Bot had a problem deploying to test February 25, 2026 01:53 Error

copy-pr-bot Bot temporarily deployed to test February 25, 2026 01:57 Inactive

copy-pr-bot Bot temporarily deployed to test February 25, 2026 07:23 Inactive

copy-pr-bot Bot temporarily deployed to test February 25, 2026 10:18 Inactive

copy-pr-bot Bot temporarily deployed to test February 25, 2026 15:37 Inactive

Merge branch 'main' into dev

0ab47fa

FDecaYed force-pushed the deyuf/dev_pull_main_260220 branch from 6531920 to 0ab47fa Compare February 26, 2026 03:31

copy-pr-bot Bot temporarily deployed to test February 26, 2026 03:34 Inactive

chtruong814 added the Run functional tests label Feb 26, 2026

ko3n1g merged commit a1a73f8 into NVIDIA:dev Feb 26, 2026
45 of 48 checks passed

Navigation Menu

Uh oh!

[dev] pull main 260220 - #3574

[dev] pull main 260220#3574
ko3n1g merged 219 commits into
NVIDIA:devfrom
FDecaYed:deyuf/dev_pull_main_260220

FDecaYed commented Feb 25, 2026

Uh oh!

copy-pr-bot Bot commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 26, 2026

Uh oh!

chtruong814 commented Feb 26, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

Uh oh!

Conversation

FDecaYed commented Feb 25, 2026

What does this PR do ?

Contribution process

Pre-checks

Code review

(Step 1): Add PR label Expert Review

(Step 2): Collect the expert reviewers reviews

(Step 3): Final Review

(Optional Step 4): Cherry-pick into release branch

Merging your PR

Uh oh!

copy-pr-bot Bot commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 25, 2026

Uh oh!

FDecaYed commented Feb 26, 2026

Uh oh!

chtruong814 commented Feb 26, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

(Step 1): Add PR label `Expert Review`