Skip to content

From NVIDIA Megatron-LM for visibility#18

Open
RaymondLi0 wants to merge 6861 commits intobigcode-project:multi-query-attentionfrom
NVIDIA:main
Open

From NVIDIA Megatron-LM for visibility#18
RaymondLi0 wants to merge 6861 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
i-riyad and others added 27 commits March 26, 2026 05:16
@i-riyad @santhnm2
Pass gracefully if token_id not found in message (#3862)
606afda 
Signed-off-by: rislam <rislam@nvidia.com>
Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>
Co-authored-by: Keshav Santhanam <ksanthanam@nvidia.com>
@santhnm2
Improve load balancing behavior for prefix cache-aware routing (#3930)
0528a40 
Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>
@sakgoyal
Refactor setup.py to use get_pybind_include (#3658)
58e0b85
@ko3n1g
build: Bump TE to 2.14 (#4025)
3758b54 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@github-actions
chore(beep boop 🤖): Bump (main) (2026-03-30)
d863b7b
@dimapihtar
fix traceback when interrupting run (#3439)
a61ce5f 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@ko3n1g
chore: update goldenvalues (#4059)
4dcd7d6 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@lvdunlin @dimapihtar
Fix TemporalAsyncCaller pin_memory lifetime in async checkpointing (#…
fc61ce5 
…2288)

Co-authored-by: lvdunlin <lvdunlin@xiaomi.com>
Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com>
@ko3n1g @claude
chore: Move to Py3.12 (#3826)
4bde3a4 
Signed-off-by: oliver könig <okoenig@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
@github-actions
Update copy-pr-bot.yaml [skip ci]
8256553
@dimapihtar @sbak5 @claude
Adding NVRx as a dependency and keeping the current code base optiona…
704c7ee 
…lly (#3899)

Signed-off-by: dimapihtar <dpihtar@gmail.com>
Co-authored-by: Seonmyeong Bak <sbak@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
@ko3n1g
chore: Bump versions
e1db321 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@ko3n1g
build: Set ENV NVTE_BUILD_NUM_PHILOX_ROUNDS=3 (#4074)
8dd65cd 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@dimapihtar
fix checkpointing conversion (#4058)
a3a7a0c 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@ko3n1g
chore: Bump versions
f09f5c9 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@ko3n1g @claude
fix(ci): replace actions/setup-python with apt-get to avoid 429 rate …
dc113cf 
…limits (#4072)

Signed-off-by: oliver könig <okoenig@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
@ko3n1g
ci: Fix package name for code-freeze workflow (#4077)
8f3bee5 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@ko3n1g @claude
chore: bump _code_freeze workflow to v0.86.0 (#4078)
1533beb 
Signed-off-by: oliver könig <okoenig@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
@janEbert
Fix checkpoint inspector (#4079)
ef2c8a0
@megnvidia
Update docs to conform to NVIDIA style guides (#4068)
fd1888b 
Signed-off-by: meg miranda <mmiranda@nvidia.com>
@santhnm2 @ko3n1g
@lvdunlin @dimapihtar @claude @github-actions
Miscellaneous inference fixes (#4030)
2b85d0a 
Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>
Signed-off-by: oliver könig <okoenig@nvidia.com>
Co-authored-by: oliver könig <okoenig@nvidia.com>
Co-authored-by: ldl <140483453+lvdunlin@users.noreply.github.com>
Co-authored-by: lvdunlin <lvdunlin@xiaomi.com>
Co-authored-by: Dmytro Pykhtar <37850217+dimapihtar@users.noreply.github.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: GitHub Actions <github-actions[bot]@users.noreply.github.com>
@CarlosGomes98
fix fine_grained_callables with fused rmsnorm residual (#4026)
15f14fc
@Wohox
[Main][feat] Support overlapping A2A Combine backprop with wgrad GEMM (…
97e36aa 
…#3795)
@github-actions
chore: rotate oncall schedule
7086b61
@wplf @cspades
Modify mfsdp default data-parallel-sharding-strategy (#3691)
3499efe 
Co-authored-by: Jinliang Li <jinliangl@eos0065.eos.clusters.nvidia.com>
Co-authored-by: Cory Ye <44509866+cspades@users.noreply.github.com>
@DAISY-gh @cspades
Fix fsdp_dtensor conversion for pretrained-only checkpoints (#3912)
1284d25 
Co-authored-by: Cory Ye <44509866+cspades@users.noreply.github.com>
@wdykas
Guard NVshmem issues (#4093)
f9a61e3 
Co-authored-by: root <root@nvl72063-T02.cm.cluster>
Co-authored-by: William Dykas <wdykas@oci-hsg-cs-001-vscode-03.cm.cluster>
yashaswikarnati and others added 30 commits April 28, 2026 01:29
@yashaswikarnati @claude
Add ColocatedBridgeCommunicator for heterogeneous TP/DP MIMO training…
42e396e 
… (NMFW-17) (#4368)

Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@HelloWorldBeginner @asolergi-nv
Fix incorrect bias display in extra_repr of Column/RowParallelLinear (#…
6fd6652 
…4330)

Co-authored-by: mhh111 <mahonghao1@huawei.com>
Co-authored-by: Antoni-Joan Solergibert <asolergibert@nvidia.com>
@joapolarbear @deepakn94 @asolergi-nv
Fix assertion logic in combined_1f1b_schedule_for_interleaved_pipelin…
c8a4bfd 
…ing (#4276)

Co-authored-by: Hanpeng Hu <haaanpeng@outlook.com>
Co-authored-by: Deepak Narayanan <deepakn94@gmail.com>
Co-authored-by: Antoni-Joan Solergibert <asolergibert@nvidia.com>
@balasaajay
ci: Fix event name reference in CI workflow condition for merge group (…
374fa85 
…#4462)
@Phlip79 @claude @ko3n1g
Add manual sync workflow from main to dev (#4165)
9c15290 
Signed-off-by: oliver könig <okoenig@nvidia.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: Philip Petrakian <>
Co-authored-by: oliver könig <okoenig@nvidia.com>
@ChenhanYu @claude @jenchen13
fix: handle list-format quant_cfg from ModelOpt PR #1094 (#4187)
9816140 
Signed-off-by: chenhany <chenhany@nvidia.com>
Signed-off-by: Chenhan Yu <chenhany@nvidia.com>
Signed-off-by: Jennifer Chen <jennifchen@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: Jennifer Chen <jennifchen@nvidia.com>
@Phlip79 @claude
ci: also add Run MBridge tests label in nightly sync workflow (#4499)
9e98259 
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@github-actions
Update copy-pr-bot.yaml [skip ci]
533dc75
@maanug-nv
[training migration] Add serialization features to config container (#…
1c4e537 
…4309)

Signed-off-by: Maanu Grover <maanug@nvidia.com>
@tdene
Fix conflict with inference graphs (#4504)
f4a49cf
@github-actions
chore: rotate oncall schedule
251c6e9
@asolergi-nv @claude
Add tools/prepare_cache.py for offline GPT dataset cache preparation (#…
c5201a0 
…4080)

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
@ko3n1g @claude
[build] fix: move mamba-ssm and causal-conv1d to optional [ssm] extra (…
cb3d5d9 
…#4517)

Signed-off-by: oliver könig <okoenig@nvidia.com>
Signed-off-by: Oliver Koenig <okoenig@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
@wdykas @claude
mamba: avoid redundant HBM reloads in causal_conv1d_update shift loop (…
4e208a8 
…#4460)

Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-authored-by: root <root@nvl72087-T13.cm.cluster>
Co-authored-by: William Dykas <wdykas@oci-hsg-cs-001-vscode-03.cm.cluster>
@tdene @mathemakitten
Standardize misc graph interface (#4485)
3f59bbb 
Co-authored-by: mathemakitten <helen.ngo14@gmail.com>
@tdene
Fix inference graph override in RL flow (#4323)
29864b2
@cspades @megnvidia
Unify and refactor Megatron-FSDP documentation. (#4418)
b23aa3f 
Signed-off-by: Cory Ye <cye@nvidia.com>
Co-authored-by: megnvidia <mmiranda@nvidia.com>
@chtruong814
Revert "ci: add base_sha to codecov/codecov-action upload step (#4445)…
51ea07e 
…" (#4526)

Signed-off-by: Charlie Truong <chtruong@nvidia.com>
@yashaswikarnati
Skills for running unit tests and working with slurm (#4502)
cfee04e
@tdene
Reorganize order of operations in inference context and text generati…
0d98cb8 
…on controller (#2929)
@balasaajay
ci: Update CI workflow conditions to include merge group handling (#4532
0c52c39 
)
@chtruong814
ci: add base_sha to codecov/codecov-action upload step (#4540)
6ba794b 
Signed-off-by: Charlie Truong <chtruong@nvidia.com>
@deepakn94 @claude
Fix release tests: remove --global-batch-size conflicting with --step…
580d53a 
…-batch-size-schedule (#4545)

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
@ko3n1g
docs: use @file-path notation for file references in skills (#4542)
77afc60 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@jenchen13 @claude
Support YAML quant recipe in PTQ and remove first/last layer modifier…
1a83320 
… code (#4503)

Signed-off-by: Jennifer Chen <jennifchen@nvidia.com>
Co-authored-by: claude[bot] <209825114+claude[bot]@users.noreply.github.com>
@tdene
Avoid nsys profile crash with CUDA graphs (#4541)
12f18da
@ko3n1g
fix(ci): add retry with backoff to approve-test-queue bot (#4559)
dcb2bd2 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@sidsingh-nvidia
New allgathervdispatcher for inference and simplify old dispatcher. (#…
bfd4574 
…4258)
@jenchen13 @claude
Fixes for modelopt examples and SFTTokenizer for transformers v5 (#4450)
83e7466 
Signed-off-by: Jennifer Chen <jennifchen@nvidia.com>
Signed-off-by: jenchen13 <jennifchen@nvidia.com>
Co-authored-by: claude[bot] <209825114+claude[bot]@users.noreply.github.com>
@github-actions
Update copy-pr-bot.yaml [skip ci]
3460bba
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.