Skip to content

Comments

Fetch from nvidia Megatron-LM#5

Open
RaymondLi0 wants to merge 6693 commits intoElementAI:load-iterfrom
NVIDIA:main
Open

Fetch from nvidia Megatron-LM#5
RaymondLi0 wants to merge 6693 commits intoElementAI:load-iterfrom
NVIDIA:main

Conversation

@RaymondLi0
Copy link

@RaymondLi0 RaymondLi0 commented Aug 3, 2022

No description provided.

maanug-nv and others added 30 commits January 28, 2026 07:13
@maanug-nv
[training migration] Add CheckpointConfig dataclass (#2431)
d4f9347 
Signed-off-by: Maanu Grover <maanug@nvidia.com>
@github-actions
chore: rotate oncall schedule
d5cac80
@maanug-nv
[training migration] Add StragglerDetectionConfig dataclass (#2435)
008926a 
Signed-off-by: Maanu Grover <maanug@nvidia.com>
@tdene
Standardize RL unit tests (#3088)
1453f94
@HaochenYuan @Phlip79 @ko3n1g
Fix for PR-2142 (#3096)
71c49b5 
Co-authored-by: Philip Petrakian <ppetrakian@nvidia.com>
Co-authored-by: oliver könig <okoenig@nvidia.com>
@Autumn1998
Use the latest hybrid-ep (#3093)
93ddc24
@dimapihtar
remove retro (#3001)
fc6969f 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@ko3n1g
ci: Mark test_compatible_with_nd_parallel as flaky (#3122)
c22615e 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@ko3n1g
build: Use merge-commit-sha for container (#3123)
a883e96 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@mathemakitten
Refactor rl_offload_kv_cache_during_training to offload KV cache to…
42986ac 
… CPU while retaining fixed virtual address (#3048)
@Phlip79
Disable Greptile status comments (#3127)
d41bf66
@github-actions
Update copy-pr-bot.yaml [skip ci]
e2ff203
@ko3n1g
ci: Add unit tests to merge queue (#3125)
50132f2 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@Phlip79
Create CodeRabbit config (#3131)
9f05aac
@chtruong814
build: Explicitly set minimum torch version to >= 2.6.0 (#3085)
f0b1cb2 
Signed-off-by: Charlie Truong <chtruong@nvidia.com>
@kwyss-nvidia @jaredcasper
Move kitchen extension file to private kitchen repository (#2779)
190f5b6 
Co-authored-by: Jared Casper <155158+jaredcasper@users.noreply.github.com>
@jon-barker
Fix RL optimizer offload (#3112)
287d2f4 
Co-authored-by: Jon Barker <jbarker@oci-hsg-cs-001-vscode-01.cm.cluster>
@ko3n1g
Revert "Fix RL optimizer offload (#3112)" (#3141)
3955c49
@AAnoosheh @ko3n1g
Revise and move KD docs (#3108)
4913c46 
Signed-off-by: Asha Anoosheh <aanoosheh@nvidia.com>
Co-authored-by: oliver könig <okoenig@nvidia.com>
@ko3n1g
build: Bump FLA (#3139)
558fdaf 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@ko3n1g
ci: Add job timeouts (#3142)
409af92 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@ko3n1g
ci: Set NODE_RANK (#3143)
f4af1bf 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@yobibyte
Multiturn rollout support prep (#2966)
0b619c2
@jon-barker
Reapply 3955c49 (#3146)
36411dd
@ko3n1g
Revert "Multiturn rollout support prep (#2966)" (#3153)
dbd8dda 
Signed-off-by: oliver könig <okoenig@nvidia.com>
@Phlip79
Fix coderabbit instructions error (#3150)
f58b6d6
@asolergi-nv
Force input ids generated by mock dataset are < vocab_size (#2945)
063624b
@asolergi-nv
Add a check to make sure we are distributing all the layers when usin…
4652e7b 
…g `--decoder-first-pipeline-num-layers` & `--decoder-last-pipeline-num-layers` (#2947)
@github-actions
Update copy-pr-bot.yaml [skip ci]
18deeff
@tdene
Automatically choose available ports in ZMQ (#2278)
67f3515
github-actions bot and others added 30 commits February 21, 2026 00:14
@github-actions
Update copy-pr-bot.yaml [skip ci]
b555baf
@sidsingh-nvidia
Renable full_iteration cuda graphs for inference. Add them for the ma…
32efeff 
…mba block. (#3250)
@arendu @rkarimimahab @seonjinn
do not add EoD (#3526)
a6d6dc6 
Signed-off-by: adithyare <adithyare@nvidia.com>
Signed-off-by: Soumye Singhal <soumyes@cw-dfw-cs-001-dc-01.cm.cluster>
Co-authored-by: Soumye Singhal <soumyes@cw-dfw-cs-001-dc-01.cm.cluster>
Co-authored-by: Rabeeh Karimi Mahabadi <rkarimimahab@nvidia.com>
Co-authored-by: Seonjin Na <sna@nvidia.com>
@github-actions
chore(beep boop 🤖): Bump (main) (2026-02-23)
7124748
@Phlip79
Do not Slack notify for draft PRs (#3536)
3d1a4ba
@github-actions
Update copy-pr-bot.yaml [skip ci]
9c6b69b
@dimapihtar
remove deprecated SampleListWebdataset (#3407)
6159399 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@dimapihtar
remove deprecated get_te_version (#3413)
4aeeca8 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@dimapihtar
remove deprecated async_grad_allreduce param (#3412)
33df24c 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@dimapihtar
remove deprecated mamba params (#3411)
eb0da52 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@dimapihtar
remove deprecated params from model parallel config (#3408)
82cbd82 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@duncanriach
Remove redundant CUDA calls in the LLaVA dataloader (#3476)
a0b9c16
@sidsingh-nvidia
Inference: Create finer grained cuda-graphs with better coverage of s…
fde3b90 
…maller batch sizes (#3527)
@ahmadki
fix: skip non-tensor optimizer state entries in distrib_optimizer sav… (
23dd639 
#3537)

Signed-off-by: Ahmad Kiswani <kiswani.ahmad@gmail.com>
@dimapihtar
remove is_unitialized & get_data_modulo_expert_parallel_group (#3414)
2ec295e 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@dimapihtar
remove deprecated TE module (#3409)
abf97eb 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@dimapihtar
remove encoder_and_decoder from enums (#3406)
7c7c9e1 
Signed-off-by: dimapihtar <dpihtar@gmail.com>
@github-actions
chore(beep boop 🤖): Bump (main) (2026-02-24)
dd39eb5
@sbak5
Add knobs to choose process groups for fully-parallel-save / load and…
cb24802 
… load-exchange-algo (#2161)
@tdene
Fix off-by-2 error in RL sequence packing (#3551)
5f5f465
@sbak5
Skip unnecessary flattening for Save / Load Planner (#3263)
a8efd34
@faradawn
Multimodal: fix model provider (#3508)
5dc98a6 
Signed-off-by: Faradawn Yang <73060648+faradawn@users.noreply.github.com>
@chtruong814
docs: Enable nightly docs publish (#3546)
76b200c 
Signed-off-by: Charlie Truong <chtruong@nvidia.com>
@nschank
Ensure type-checker understands use of Submodules in unit tests (#3425)
f721069
@nschank
Use copy_signature to preserve typing of pass-through methods (#3419)
782e54b
@nschank
Ensure type-checker understands use of Submodules in MTP (#3308)
3597312
@santhnm2
Add mxfp8 quantization for inference linear layers (#3447)
44e27d0 
Signed-off-by: Keshav Santhanam <ksanthanam@nvidia.com>
@sbak5
Add single-process checkpoint save to avoid forked multiprocessing (#…
2cac78b 
…3424)
@mkhona-nvidia
Fixed fp32 residuals (#3504)
08857d9 
Signed-off-by: mikail <mkhona@nvidia.com>
@maanug-nv
Move config src files into a dedicated dir (#3570)
b17248a 
Signed-off-by: Maanu Grover <maanug@nvidia.com>
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.