Skip to content

Pull requests: NVIDIA-NeMo/RL

New pull request New
107 Open 1,119 Closed
107 Open 1,119 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

feat: use ray.put to reduce overhead of repeated arguments super-v3
#1944 opened Feb 14, 2026 by guyueh1 Loading…
4 tasks
3
feat: Add HybridEP support for MoE expert parallelism
#1942 opened Feb 13, 2026 by seonjinn Loading…
4 tasks
6
Shaunak/vllm specdec metrics community-request documentation Improvements or additions to documentation
#1941 opened Feb 13, 2026 by shaunjoshi Loading…
4 tasks
11
Remove do_not_average_loss; undo Megatron loss averaging in RL code
#1940 opened Feb 13, 2026 by yaoyu-33 Loading…
1
build: Update dockerfile to support Nsight install on arm platforms CI:L2 Run doctests, unit tests, functional tests, and convergence tests
#1939 opened Feb 12, 2026 by ananthsub Loading…
4 tasks
1
feat: Support top-p and top-k
#1938 opened Feb 12, 2026 by zhandaz Loading…
2 of 4 tasks
13
Fix mcore inference
#1931 opened Feb 12, 2026 by shanmugamr1992 Loading…
4 tasks
fix: Fix device mismatch when DPO runs validation at start with CPU offload (Nemotron MoE) CI:L1 Run doctests, unit tests, and functional tests
#1930 opened Feb 12, 2026 by RayenTian • Draft
4 tasks
feat: add draft model support community-request documentation Improvements or additions to documentation needs-follow-up Issue needs follow-up
#1921 opened Feb 10, 2026 by shaunjoshi • Draft
4 tasks
6
refactor: refactor loss function
#1920 opened Feb 10, 2026 by yuki-97 • Draft
7
perf: Fuse sequence packing for loss function
#1904 opened Feb 10, 2026 by nujoug Loading…
7
Add Word-guess NemoGym GRPO training
#1903 opened Feb 9, 2026 by yyu22 • Draft
1
test: Add script for nemotron test CI:L0 Run doctests and unit tests super-v3
#1901 opened Feb 9, 2026 by guyueh1 Loading…
4 tasks
@guyueh1
1
feat: ProRLv2 - add seq-mask-tis truncated importance sampling type CI:L1 Run doctests, unit tests, and functional tests documentation Improvements or additions to documentation
#1899 opened Feb 9, 2026 by hijkzzz Loading…
@hijkzzz
13
perf: Reduce memory footprint for ChunkedDistribuedLogProb
#1895 opened Feb 6, 2026 by nujoug Loading…
1
@nujoug
6
Draft: Use Megatron-Bridge recipes for megatron_cfg.
#1894 opened Feb 6, 2026 by sfawzy-nv • Draft
4 tasks
9
feat: skip logprob and reference logprob computation under certain conditions CI:L1 Run doctests, unit tests, and functional tests deepseek Related to deepseek 671b super-v3
#1891 opened Feb 6, 2026 by guyueh1 Loading…
4 tasks
@guyueh1
8
feat: Megatron LoRA GRPO w/ Weight Merging
#1889 opened Feb 5, 2026 by vadam5 Loading…
4 tasks
1
10
feat: MXFP8 rollout support super-v3
#1887 opened Feb 5, 2026 by guyueh1 • Draft
4 tasks
@guyueh1
feat: Add perfetto tracing for async GRPO training
#1876 opened Feb 4, 2026 by gspschmid Loading…
4 tasks
12
feat: add worker initialization timing collection CI:L0 Run doctests and unit tests CI:L1 Run doctests, unit tests, and functional tests
#1873 opened Feb 4, 2026 by yashaswikarnati Loading…
4 tasks
12
chore: bump torch 2.9.1, vllm 0.15 sglang 0.5.8, ray 2.53 dependencies Pull requests that update a dependency file
#1871 opened Feb 3, 2026 by terrykong Loading…
4 tasks
3
feat: add fault injection utilities for testing fault tolerance CI:L0 Run doctests and unit tests CI:L1 Run doctests, unit tests, and functional tests super-v3
#1868 opened Feb 3, 2026 by yashaswikarnati Loading…
4 tasks
@yashaswikarnati
17
fix: Remove obsolete MoE aux loss validation assertion community-request
#1852 opened Jan 30, 2026 by mingmao-inflection Loading…
4 tasks done
1
3
Mdp
#1849 opened Jan 29, 2026 by shanmugamr1992 Loading…
4 tasks
18
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.