Skip to content

Pull requests: NVIDIA-NeMo/RL

New pull request New
241 Open 1,456 Closed
241 Open 1,456 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

beep boop 🤖: Bumping NeMo-RL to v0.6.1 CI Relating to CI
#2549 opened May 22, 2026 by nemo-automation-bot Bot Loading…
1
beep boop 🤖: Bumping NeMo-RL to v0.6.1
#2548 opened May 22, 2026 by nemo-automation-bot Bot Loading…
1
beep boop 🤖: Bumping NeMo-RL to v0.6.1
#2547 opened May 22, 2026 by nemo-automation-bot Bot Loading…
1
ci: pin FW-CI-templates to NVIDIA-NeMo/FW-CI-templates#480 CI:docs Run doctest CI Relating to CI
#2546 opened May 22, 2026 by ko3n1g Contributor Loading…
16
fix: set TIS type for Automodel GRPO release recipes
#2544 opened May 22, 2026 by zpqiu Contributor • Draft
4 tasks done
1
feat: make only_unmask_final configurable in SFT
#2543 opened May 22, 2026 by ashors1 Contributor Loading…
4 tasks
2
refactor: avoid requiring Megatron for model_utils import
#2540 opened May 21, 2026 by alexbowe • Draft
3 of 4 tasks
1
refactor: defer decord2 initialization for audio fallback
#2539 opened May 21, 2026 by alexbowe • Draft
1
feat: SDPO community-request
#2538 opened May 21, 2026 by bogdansalyp Contributor • Draft
4 tasks
1
ci: switch sglang to prebuilt PyPI wheels (v0.5.11) CI:L1 Run doctests, unit tests, and functional tests
#2535 opened May 20, 2026 by kajalj22 Contributor • Draft
4 tasks
3
[feat:] Add CISPO loss community-request Documentation Improvements or additions to documentation waiting-on-maintainers Waiting on maintainers to respond
#2531 opened May 19, 2026 by pengdurice Contributor Loading…
1 of 4 tasks
1
feat: PPO with MCore CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version)
#2530 opened May 19, 2026 by bg51717 Contributor Loading…
4 tasks done
1
2
feat: add AsyncNemoGymRolloutManager for gym per-prompt rollouts CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version)
#2528 opened May 19, 2026 by yuki-97 Contributor Loading…
1 task done
3
refactor(distillation): migrate DistillationConfig, DistillationSaveS… CI:L1 Run doctests, unit tests, and functional tests
#2527 opened May 19, 2026 by NolenLiang Contributor Loading…
4 tasks
8
refactor(rm): migrate RMConfig, RMSaveState, RMValMetrics to BaseModel CI:L1 Run doctests, unit tests, and functional tests
#2526 opened May 19, 2026 by NolenLiang Contributor Loading…
4 tasks
8
refactor(sft): migrate SFTConfig, SFTSaveState to BaseModel CI:L1 Run doctests, unit tests, and functional tests
#2525 opened May 19, 2026 by NolenLiang Contributor Loading…
4 tasks
6
refactor(dpo): migrate DPOConfig, DPOSaveState, DPOValMetrics to Base… CI:L1 Run doctests, unit tests, and functional tests
#2524 opened May 19, 2026 by NolenLiang Contributor Loading…
4 tasks
10
ci: build container on Azure for main builds CI Relating to CI
#2521 opened May 18, 2026 by kajalj22 Contributor • Draft
1 of 4 tasks
5
refactor(loss): migrate DPOLossConfig, DistillationLossConfig, DraftC… CI:L1 Run doctests, unit tests, and functional tests
#2520 opened May 18, 2026 by NolenLiang Contributor Loading…
4 tasks
13
refactor(grpo): migrate TypedDict configs to pydantic BaseModel CI:L1 Run doctests, unit tests, and functional tests
#2518 opened May 18, 2026 by NolenLiang Contributor Loading…
4 tasks
16
feat: fix the vLLM DP path CI:L2 Run doctests, unit tests, functional tests, and convergence tests
#2517 opened May 18, 2026 by guyueh1 Contributor Loading…
4 tasks
1
14
feat(sft): make only_unmask_final configurable in SFTConfig CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version)
#2516 opened May 17, 2026 by yuki-97 Contributor Loading…
1 task done
1
3
fix: fix preserving dataset merge CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version)
#2515 opened May 17, 2026 by yuki-97 Contributor Loading…
1 task done
1
3
fix(nemo-gym): clamp max_new_tokens to prompt + output <= max_model_len CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version)
#2514 opened May 17, 2026 by yuki-97 Contributor Loading…
2 tasks done
2
fix(vllm): serialise AsyncMPClient input_socket sends to prevent zmq race CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version) community-request
#2513 opened May 17, 2026 by kaloyan-inherent Loading…
1
10
ProTip! Add no:assignee to see everything that’s not assigned.