Skip to content

Pull requests: NVIDIA-NeMo/RL

New pull request New
332 Open 1,955 Closed
332 Open 1,955 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(xtoken): replace DP token alignment with offset-based cluster alignment
#3286 opened Jul 19, 2026 by avenkateshha Contributor Loading…
2 tasks done
1
fix: add missing closing brace and default value for extra_head in conf.py (closes #2816) community-request Documentation Improvements or additions to documentation
#3285 opened Jul 19, 2026 by botbikamordehai2-sketch Loading…
1
1
fix(envs): stamp per-attempt prompt_id on nemo_gym rollouts to isolate verify cohorts
#3284 opened Jul 19, 2026 by bg51717 Contributor • Draft
2 of 4 tasks
1
perf(nemo_gym): hand tokenizer to NemoGym actor once at spinup
#3283 opened Jul 19, 2026 by bg51717 Contributor • Draft
3 of 4 tasks
1
fix: fix single controller grad_norm CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version)
#3282 opened Jul 19, 2026 by yuki-97 Contributor Loading…
2
fix: plumb fp8_recipe into OptimizerConfig for fp8_param runs
#3281 opened Jul 18, 2026 by seonjinn Contributor Loading…
1
build: bump vllm 0.20.0 → 0.25.1 CI:L1 Run doctests, unit tests, and functional tests
#3280 opened Jul 18, 2026 by terrykong Collaborator • Draft
15
perf: increase Qwen3-235B async rollout GPU memory utilization on H100 CI:L2 Run doctests, unit tests, functional tests, and convergence tests
#3278 opened Jul 18, 2026 by humairafirdowse18 Loading…
4 tasks done
1
feat: ultra recipes for gb200 Documentation Improvements or additions to documentation
#3276 opened Jul 17, 2026 by macandro96 Contributor • Draft
4 tasks
1
fix(data): warn when behavioral dataset config keys are silently ignored community-request waiting-on-maintainers Waiting on maintainers to respond
#3271 opened Jul 17, 2026 by tianyi-zhang-02 Contributor Loading…
3 tasks done
1
1
ci: AUT-895 gate GB200 tests on DISABLE_GB200_TESTS variable CI:L1 Run doctests, unit tests, and functional tests CI Relating to CI
#3269 opened Jul 17, 2026 by svcnemo-autobot Loading…
3
feat(telemetry): instrument NeMo-RL with OpenTelemetry via nemo-lens Documentation Improvements or additions to documentation
#3268 opened Jul 17, 2026 by ahmadki Member • Draft
4 tasks done
4
feat(sc): NeMo-Gym path
#3267 opened Jul 17, 2026 by yuki-97 Contributor • Draft
@RayenTian
1
feat(sc): setup + entrypoint CI Relating to CI
#3266 opened Jul 17, 2026 by yuki-97 Contributor • Draft
@RayenTian
1
feat(sc): abort stale in-flight rollouts early under over_sampling
#3263 opened Jul 17, 2026 by RayenTian Contributor • Draft
4 tasks
1
feat(ppo): support non-colocated generation CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version) Documentation Improvements or additions to documentation
#3262 opened Jul 17, 2026 by bg51717 Contributor Loading…
4 tasks done
4
fix(skills): make review-pr frontmatter Open-Plugin-valid (string fields)
#3261 opened Jul 17, 2026 by ko3n1g Contributor • Draft
1
fix(skills): quote review-pr argument-hint so SKILL.md frontmatter parses
#3260 opened Jul 17, 2026 by ko3n1g Contributor Loading…
1
feat(weight-sync): add ModelExpress trainer publishing community-request Documentation Improvements or additions to documentation
#3259 opened Jul 17, 2026 by KavinKrishnan Loading…
5 of 6 tasks
3
feat: Combined RLVR + MOPD implementation + recipe
#3256 opened Jul 16, 2026 by rohitrango Contributor • Draft
1 of 4 tasks
1
feat(vlm): support configurable custom reward functions community-request Documentation Improvements or additions to documentation waiting-on-maintainers Waiting on maintainers to respond
#3250 opened Jul 16, 2026 by abhinav-bellapu Loading…
4 of 5 tasks
1
2
feat(vllm): expose Responses API from async workers community-request Documentation Improvements or additions to documentation waiting-on-maintainers Waiting on maintainers to respond
#3249 opened Jul 16, 2026 by abhinav-bellapu Loading…
4 of 5 tasks
2
fix(grpo): honor max_num_epochs in async training community-request Documentation Improvements or additions to documentation waiting-on-maintainers Waiting on maintainers to respond
#3248 opened Jul 16, 2026 by abhinav-bellapu Loading…
4 of 5 tasks
2
docs: warn about ray.sub shared-venv drift and pre-sync fix Documentation Improvements or additions to documentation
#3247 opened Jul 16, 2026 by achartier Contributor Loading…
2 of 4 tasks
2
Feat/nemo gym eval telemetry Documentation Improvements or additions to documentation
#3243 opened Jul 16, 2026 by joyang-nv Member Loading…
4 tasks
1
ProTip! Exclude everything labeled bug with -label:bug.