Skip to content

From NVIDIA Megatron-LM for visibility#18

Open
RaymondLi0 wants to merge 6919 commits intobigcode-project:multi-query-attentionfrom
NVIDIA:main
Open

From NVIDIA Megatron-LM for visibility#18
RaymondLi0 wants to merge 6919 commits intobigcode-project:multi-query-attentionfrom
NVIDIA:main

Conversation

@RaymondLi0
Copy link
Copy Markdown
Collaborator

@RaymondLi0 RaymondLi0 commented Jan 24, 2023

No description provided.

@RaymondLi0 RaymondLi0 changed the base branch from multi-query-attention to before-merge June 20, 2023 20:12
@RaymondLi0 RaymondLi0 changed the base branch from before-merge to multi-query-attention June 20, 2023 20:12
Phlip79 and others added 27 commits April 6, 2026 17:23
@Phlip79 @claude
Claude: add respond-to-issue skill (#4141)
7d536c0 
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
@skyw
Fix muon getter backward compatability (#4157)
6652d57 
Signed-off-by: Hao Wu <skyw@nvidia.com>
@megnvidia
Audit of user guide (#4098)
97cd326 
Signed-off-by: meg miranda <mmiranda@nvidia.com>
@yezhengmao1 @dimapihtar
Fix RerunStateMachine crash (`TypeError: 'NoneType' object is not s…
0b5e3ae 
…ubscriptable`) by not saving a checkpoint after a transient NaN / Inf (#3981)

Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com>
@nschank
Preserve type of decorated methods/classes (#4062)
fa5103c
@skyw
update muon test case to use new interface (#4163)
eba2eaf 
Signed-off-by: Hao Wu <skyw@nvidia.com>
@shjwudp @cspades @conver334
[M-FSDP] Fix Tensor Parallel mode detection (#3191)
8cbc45b 
Signed-off-by: Cory Ye <cye@nvidia.com>
Co-authored-by: Cory Ye <cye@nvidia.com>
Co-authored-by: conver334 <conver334@gmail.com>
@faradawn
fix: remove weights_only=False for multimodal example (#4104)
8e8aff6 
Signed-off-by: Faradawn Yang <73060648+faradawn@users.noreply.github.com>
@mathemakitten
Cudagraphs: Fix sequence packing segfault more generally (#4162)
52150ab
@github-actions
Update copy-pr-bot.yaml [skip ci]
70a7f69
@santhnm2
Make MTP work with materialize_only_last_token_logits (#4166)
c146305 
Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>
@santhnm2
Add unit test for Mamba EP inference (eager fallback with mixed CUDA …
8cf6b35 
…graphs) (#4085)

Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>
@github-actions
chore: rotate oncall schedule
2368a2e
@dimapihtar
update docs in respect to async changes (#4177)
cc025f8 
Signed-off-by: dimapihtar <dpykhtar@nvidia.com>
@dimapihtar
update checkpointing docs in respect to async changes (#4208)
1341d8c 
Signed-off-by: dimapihtar <dpykhtar@nvidia.com>
@ko3n1g @claude
chore: improve build-and-test skill with trigger rules and dependency…
e40feed 
… workflow (#4199)

Signed-off-by: oliver könig <okoenig@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
@skyw
Fix layerwise optimizer with expt_dp_size=1 and contention with ele…
51bcf14 
…ment-wise distributed optimizer (#4138)

Signed-off-by: Hao Wu <skyw@nvidia.com>
@ko3n1g @claude
ci: add --cluster-a100/h100/gb200 args to trigger_internal_ci.py (#4195)
50bafa0 
Signed-off-by: oliver könig <okoenig@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
@chtruong814
ci: Update golden values for nightly tests (#4215)
41595d0 
Signed-off-by: Charlie Truong <chtruong@nvidia.com>
@github-actions
Update copy-pr-bot.yaml [skip ci]
a614df4
@skyw
rename async_allgather to overlap_param_gather (#4217)
1b425ab 
Signed-off-by: Hao Wu <skyw@nvidia.com>
@Phlip79 @claude
Fix Slack sync for users with GitHub email privacy enabled (#4220)
8f72c54 
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
@santhnm2
Miscellaneous MTP inference fixes (#4191)
980211a 
Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>
@mathemakitten
Move inference guards out of arguments.py (#4210)
be09bb6
@fanshiqing
Fix: enable fine-grained activation offloading for Mamba model. (#4173)
92d5c1f
@dimapihtar
bump NVRx (#4178)
b5354a8 
Signed-off-by: dimapihtar <dpykhtar@nvidia.com>
@maanug-nv
Update tokenizer args for Nemotron3 release config (#4239)
09312c8 
Signed-off-by: Maanu Grover <maanug@nvidia.com>
wdykas and others added 30 commits May 5, 2026 17:36
@wdykas
Fix Hang in tests (#4575)
ae65776
@Phlip79 @claude
Single commit for main2dev nightly (#4614)
b819ac7 
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@dimapihtar
convert tokenizer args to config (#4406)
40d024b 
Signed-off-by: dimapihtar <dpykhtar@nvidia.com>
@wdykas @sidsingh-nvidia @claude
Siddharth/fix ep sync (#4607)
190c833 
Co-authored-by: Siddharth Singh <sidsingh@nvidia.com>
Co-authored-by: root <root@nvl72098-T11.cm.cluster>
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-authored-by: root <root@nvl72078-T18.cm.cluster>
Co-authored-by: William Dykas <wdykas@oci-hsg-cs-001-vscode-03.cm.cluster>
Co-authored-by: root <root@nvl72102-T05.cm.cluster>
@megnvidia
mmiranda working on another set of broken links (#4534)
6e5fb47 
Signed-off-by: meg miranda <mmiranda@nvidia.com>
@deepakn94 @claude
Fix gradient corruption with layerwise param all-gather overlap (#4609)
b25a76e 
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
@ko3n1g
test: mark TestFusedApplyMLARope::test_forward_backward_for_q flaky_i…
c325855 
…n_dev (#4639)

Signed-off-by: oliver könig <okoenig@nvidia.com>
@github-actions
chore: rotate oncall schedule
fd443f2
@dimapihtar
remove legacy GPT code (#4322)
39ec5eb 
Signed-off-by: dimapihtar <dpykhtar@nvidia.com>
@balasaajay
ci: introduce L-tier scope vocabulary via parser (#4625)
008c3e9
@sidsingh-nvidia
Inference: Tune vLLM grouped gemm, moe_sum kernel, and enable shared …
20f0936 
…expert overlap in latent MoEs (#4603)
@tdene
Fix crash involving evicted requests and tpot (#4645)
5c7968a
@dimapihtar
remove legacy tranformer and modules (#4207)
8de8238 
Signed-off-by: dimapihtar <dpykhtar@nvidia.com>
@balasaajay
chore: Update Docker image version to 26.04-py3 (#4611)
5fe3f06
@mathemakitten
Propagate errors for failed inference requests (#4679)
387ba42
@mathemakitten
Inference: Cache input + position ID views (#4634)
1df264c
@chtruong814
ci: Update Gitlab base image to 26.04 pytorch (#4688)
7cdf652 
Signed-off-by: Charlie Truong <chtruong@nvidia.com>
@deepakn94 @claude
Add periodic GPU sniff tests to detect hardware stragglers (#4662)
932d9ee 
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
@chtruong814
ci: Bump GHA versions (#4606)
8079f37 
Signed-off-by: Charlie Truong <chtruong@nvidia.com>
@ko3n1g
build: widen flashinfer-python pin to <0.7.0 (#4700)
a7bd379 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@Shreyas-S-809
Add Python-side guardrail for HybridEP InfiniBand limit and rename se…
a08e259 
…q_len (#4094)
@janEbert
Switch oncall (#4702)
dfd4a1a
@balasaajay
Update golden values for various functional tests (#4703)
4933b24
@balasaajay
chore: Update golden values for various functional tests (#4706)
31fbdf0
@ko3n1g
build: upgrade mamba-ssm to 2.3.2.post1, causal-conv1d to 1.6.2.post1 (…
708c6f5 
…#4712)

Signed-off-by: oliver könig <okoenig@nvidia.com>
@ko3n1g
ci: replace uuidgen with /proc/sys/kernel/random/uuid (#4714)
1256299 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@ko3n1g
chore(codeowners): add megatron/inference/ ownership (#4704)
7106fe9 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@nschank @asolergi-nv
Create a Protocol for the MLP layer of TransformerLayer (#3435)
5e31514 
Co-authored-by: Antoni-Joan Solergibert <asolergibert@nvidia.com>
@ko3n1g
Revert "Add Python-side guardrail for HybridEP InfiniBand limit and r…
a2ec5c1 
…ename seq_len (#4094)" (#4718)

Signed-off-by: oliver könig <okoenig@nvidia.com>
@github-actions
chore(beep boop 🤖): Bump (main) (2026-05-11)
e93755e
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.