-
Notifications
You must be signed in to change notification settings - Fork 357
Pull requests: NVIDIA-NeMo/Megatron-Bridge
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
cp:
fix NemotronLabsDiffusionBridge inference bug (4179) into r0.5.0
cherry-pick
Run CICD
#4250
opened Jun 9, 2026 by
svcnvidia-nemo-ci
Contributor
Loading…
Enable Qwen3 GB200/GB300/GB300 MXFP8 CuTeDSL Fusion and Full-Iter CUDA Graphs for 26.06
area:perf
Performance optimizations and benchmarking
feature
New capabilities, enhancements, or enablement work
needs-more-tests
Requires additional L0 and L1 test coverage before merge
waiting-on-customer
Waiting on the original author to respond
#4249
opened Jun 9, 2026 by
rhmukundan
Contributor
Loading…
cp: Model implementations and HF bridge logic
cherry-pick
docs
Documentation-only updates or documentation debt
docs-only
With great power comes great responsibility.
needs-review
PR is ready for code review and waiting on a reviewer
Run CICD
Add Qwen3.5-VL MegatronMIMO non-colocated SFT tutorial (4239) into r0.5.0
area:model
#4248
opened Jun 9, 2026 by
svcnvidia-nemo-ci
Contributor
Loading…
cp: Model implementations and HF bridge logic
bug
Something isn't working
cherry-pick
needs-more-tests
Requires additional L0 and L1 test coverage before merge
needs-review
PR is ready for code review and waiting on a reviewer
Run CICD
[model] fix: guard GLM fused expert detection (4212) into r0.5.0
area:model
#4247
opened Jun 9, 2026 by
svcnvidia-nemo-ci
Contributor
Loading…
Cherry pick 4220 r0.5.0
area:model
Model implementations and HF bridge logic
bug
Something isn't working
needs-more-tests
Requires additional L0 and L1 test coverage before merge
needs-review
PR is ready for code review and waiting on a reviewer
#4244
opened Jun 9, 2026 by
suiyoubi
Contributor
Loading…
5 tasks
[build] chore: update pinned MCore dev commit
area:build
Dependencies, packaging, images, and environment setup
ci
CI, automation, test queue, or workflow infrastructure work
needs-review
PR is ready for code review and waiting on a reviewer
#4243
opened Jun 9, 2026 by
cuichenx
Contributor
Loading…
[build] chore: update pinned MCore dev commit
area:build
Dependencies, packaging, images, and environment setup
ci
CI, automation, test queue, or workflow infrastructure work
needs-review
PR is ready for code review and waiting on a reviewer
#4242
opened Jun 9, 2026 by
cuichenx
Contributor
Loading…
cp: Performance optimizations and benchmarking
cherry-pick
feature
New capabilities, enhancements, or enablement work
needs-review
PR is ready for code review and waiting on a reviewer
Run CICD
full-iter CG for GPT-OSS 120B (4173) into r0.5.0
area:perf
#4241
opened Jun 9, 2026 by
svcnvidia-nemo-ci
Contributor
Loading…
[diffusion] fix: make DFM module
bug
Something isn't working
community-request
needs-review
PR is ready for code review and waiting on a reviewer
select_samples_to_pack shuffle deterministic across resume
area:diffusion
#4237
opened Jun 9, 2026 by
nayopu
Loading…
2 tasks done
feat(ckpt): complete Energon dataloader checkpointing save/restore
area:data
Dataset builders, preprocessing, and samplers
community-request
feature
New capabilities, enhancements, or enablement work
needs-review
PR is ready for code review and waiting on a reviewer
#4235
opened Jun 9, 2026 by
rob-luke
Contributor
Loading…
2 of 5 tasks
chore(beep boop 🤖): Bump
uv.lock (main, mcore-dev) (2026-06-09)
full-test-suite
#4232
opened Jun 9, 2026 by
svcnvidia-nemo-ci
Contributor
Loading…
build(deps): exclude CVE codecs (av/imageio/imageio-ffmpeg) from the image
area:build
Dependencies, packaging, images, and environment setup
ci
CI, automation, test queue, or workflow infrastructure work
full-test-suite
needs-review
PR is ready for code review and waiting on a reviewer
r0.5.0
Auto-cherrypick to release branch. Apply before merge; cherrypick happens after merge.
#4229
opened Jun 9, 2026 by
ko3n1g
Contributor
Loading…
feat(recipe): DSV3 GB200 MXFP8 full-iter CG recipe
area:perf
Performance optimizations and benchmarking
feature
New capabilities, enhancements, or enablement work
needs-review
PR is ready for code review and waiting on a reviewer
#4226
opened Jun 9, 2026 by
dingqingy-nv
Contributor
Loading…
3 of 4 tasks
cp: Training loop, callbacks, and runtime integration
bug
Something isn't working
cherry-pick
needs-review
PR is ready for code review and waiting on a reviewer
Run CICD
[training] fix: Route batches to standalone MTP stages (4208) into r0.5.0
area:training
#4225
opened Jun 9, 2026 by
svcnvidia-nemo-ci
Contributor
Loading…
feat(scripts): add Ultra script prerequisites
area:data
Dataset builders, preprocessing, and samplers
feature
New capabilities, enhancements, or enablement work
waiting-on-customer
Waiting on the original author to respond
#4223
opened Jun 9, 2026 by
cuichenx
Contributor
Loading…
fix(checkpointing): include optimizer scaffold while loading
area:ckpt
Checkpoint conversion, loading, export, and save paths
bug
Something isn't working
needs-review
PR is ready for code review and waiting on a reviewer
#4222
opened Jun 9, 2026 by
cuichenx
Contributor
Loading…
feat(conversion): support distributed adapter export
area:peft
Parameter-efficient fine-tuning (LoRA, adapters)
feature
New capabilities, enhancements, or enablement work
waiting-on-customer
Waiting on the original author to respond
#4221
opened Jun 9, 2026 by
cuichenx
Contributor
Loading…
docs(data): use energon prepare for valor32k-avqa dataset build
area:data
Dataset builders, preprocessing, and samplers
docs
Documentation-only updates or documentation debt
docs-only
With great power comes great responsibility.
ready-to-merge
PR is approved, current, and only waiting for CI to pass before merge
#4219
opened Jun 9, 2026 by
cuichenx
Contributor
Loading…
[peft] Support multi-lora for Megatron models
area:peft
Parameter-efficient fine-tuning (LoRA, adapters)
community-request
feature
New capabilities, enhancements, or enablement work
needs-review
PR is ready for code review and waiting on a reviewer
#4218
opened Jun 8, 2026 by
mathewjhan
Loading…
5 tasks
feat(training): wire FGO optimizer-state offload in train loop
area:training
Training loop, callbacks, and runtime integration
blocked
Work cannot move forward until an external dependency is cleared
feature
New capabilities, enhancements, or enablement work
#4217
opened Jun 8, 2026 by
dingqingy-nv
Contributor
Loading…
1 of 2 tasks
[DO NOT MERGE] Enable HybridEP for THD multimodal training
area:training
Training loop, callbacks, and runtime integration
blocked
Work cannot move forward until an external dependency is cleared
feature
New capabilities, enhancements, or enablement work
needs-more-tests
Requires additional L0 and L1 test coverage before merge
#4213
opened Jun 8, 2026 by
zhongbozhu
Contributor
•
Draft
5 tasks
Enable Qwen3 GB200/GB300 MXFP8 CuTeDSL Fusion and Full-Iter CUDA Graphs
area:perf
Performance optimizations and benchmarking
feature
New capabilities, enhancements, or enablement work
needs-more-tests
Requires additional L0 and L1 test coverage before merge
waiting-on-customer
Waiting on the original author to respond
#4211
opened Jun 8, 2026 by
rhmukundan
Contributor
Loading…
VLM Energon Improvement for Nemotron Omni
area:data
Dataset builders, preprocessing, and samplers
feature
New capabilities, enhancements, or enablement work
needs-more-tests
Requires additional L0 and L1 test coverage before merge
needs-review
PR is ready for code review and waiting on a reviewer
#4209
opened Jun 8, 2026 by
huvunvidia
Contributor
Loading…
5 tasks
Previous Next
ProTip!
Filter pull requests by the default branch with base:main.