Skip to content

Pull requests: NVIDIA-NeMo/RL

New pull request New
153 Open 1,254 Closed
153 Open 1,254 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

chore: bump Megatron-Bridge to latest main (7110a96) documentation Improvements or additions to documentation
#2223 opened Apr 7, 2026 by yuki-97 • Draft
5
Full Dynamo integration
#2222 opened Apr 6, 2026 by jthomson04 • Draft
4 tasks
1
chore: upgrade Python from 3.12 to 3.13 CI:L1 Run doctests, unit tests, and functional tests documentation Improvements or additions to documentation
#2220 opened Apr 6, 2026 by kajalj22 • Draft
3 tasks
10
Fix preserving dataset merge for SFT community-request needs-follow-up Issue needs follow-up
#2218 opened Apr 6, 2026 by Bungmint Loading…
1
feat: add SDPO algorithm community-request
#2217 opened Apr 6, 2026 by celineltan Loading…
1 of 4 tasks
1
fix: handle PreservingDataset in dataset concatenation community-request needs-follow-up Issue needs follow-up
#2213 opened Apr 4, 2026 by RudimentaryChef Loading…
1 of 2 tasks
1
1
Fix under-utilization issue of run_async_multi_turn_rollout
#2211 opened Apr 3, 2026 by zhongbozhu • Draft
4 tasks
1
fix: convert_dcp_to_hf.py correctly loads the saved tokenizer from DCP if detected
#2210 opened Apr 3, 2026 by trias702 Loading…
1 of 3 tasks
1
Preserve batched env evaluation in async validation rollouts community-request
#2209 opened Apr 3, 2026 by taivu1998 • Draft
1
Add timeout and memory limits to CodeEnvironment execution community-request documentation Improvements or additions to documentation
#2208 opened Apr 3, 2026 by taivu1998 • Draft
1
1
Handle invalid environment rewards explicitly across environments, rollouts, and GRPO community-request
#2207 opened Apr 3, 2026 by taivu1998 • Draft
1
Add startup resume_if_exists handling for checkpoint reuse community-request documentation Improvements or additions to documentation
#2206 opened Apr 3, 2026 by taivu1998 • Draft
1
1
chore(beep boop 🤖): bump FW-CI-templates workflow pins to v0.88.0 CI Relating to CI
#2205 opened Apr 3, 2026 by svcnvidia-nemo-ci Loading…
1 task
1
fix: NVML memory query fallback for DGX Spark community-request needs-follow-up Issue needs follow-up
#2203 opened Apr 3, 2026 by dbuos Loading…
1
feat: Support aux loss normalization in RL SFT
#2194 opened Apr 2, 2026 by pthombre • Draft
4 tasks
1
docs: fix typos, grammar, and table issues from QA documentation Improvements or additions to documentation
#2193 opened Apr 2, 2026 by anwithk Loading…
4 tasks
1
fix: fix gb200 release/performance
#2189 opened Apr 2, 2026 by yuki-97 • Draft
1
feat: CISPO implementation
#2187 opened Apr 1, 2026 by slikhite-1 Loading…
4 tasks
6
fix: fix h100 release/performance
#2184 opened Apr 1, 2026 by yuki-97 • Draft
1
fix: skip loading reference model when KL penalty is zero
#2178 opened Mar 31, 2026 by yfw Loading…
4 tasks
1
2
fix: skip prev_logprobs computation when force_on_policy_ratio is true
#2177 opened Mar 30, 2026 by yfw • Draft
4 tasks
1
fix: use prompt token length for advantage group extraction CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version) super-v3
#2176 opened Mar 30, 2026 by yfw Loading…
4 tasks
1
2
fix: aggregate rollout metrics by semantic type (min/max/sum/mean) super-v3
#2175 opened Mar 30, 2026 by yfw Loading…
4 tasks
1
2
fix: skip_reference_policy_logprobs_calculation=true crashes training
#2174 opened Mar 30, 2026 by ShriyaRishab Loading…
4 tasks
1
1
feat: Merge megatron checkpoints with lora adapters and convert to HF format CI:Lfast Runs a fast test suite and re-use nightly `main` container (but sync dependencies to PRs version) community-request documentation Improvements or additions to documentation needs-follow-up Issue needs follow-up
#2173 opened Mar 30, 2026 by pengdurice Loading…
4 tasks done
14
ProTip! What’s not been updated in a month: updated:<2026-03-07.