Skip to content

Pull requests: NVIDIA-NeMo/RL

New pull request New
57 Open 703 Closed
57 Open 703 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

fix: Fix checkpoint overriding CI:L1 Run doctests, unit tests, and functional tests
#1255 opened Oct 2, 2025 by terrykong Loading…
4 tasks
1
4
feat: Add deepseek flops tracker
#1250 opened Oct 2, 2025 by guyueh1 Loading…
4 tasks
6
feat: add valid_tokens_per_sec metric and total_valid_tokens to save state CI:L1 Run doctests, unit tests, and functional tests r0.4.0
#1249 opened Oct 2, 2025 by terrykong Loading…
15
feat: Using mcore cpu optimizer CI:L1 Run doctests, unit tests, and functional tests
#1242 opened Oct 1, 2025 by guyueh1 Loading…
4 tasks
1
12
perf: Add a field in megatron_cfg to enable bias_activation_fusion CI:L0 Run doctests and unit tests
#1241 opened Oct 1, 2025 by katec846 Loading…
4 tasks
@katec846
6
docs: add missing async_grpo.enabled flag to configuration asyncRL CI:docs Run doctest documentation Improvements or additions to documentation r0.4.0
#1237 opened Sep 30, 2025 by youngeunkwon0405 Loading…
4 tasks
@youngeunkwon0405
4
feat: more numerically stable qwen custom plan
#1235 opened Sep 30, 2025 by terrykong Loading…
1
2
build: Fix ngc pytorch build with deep-ep
#1234 opened Sep 30, 2025 by chtruong814 Loading…
4 tasks
7
chore: Log the initial training master config
#1232 opened Sep 29, 2025 by pjin-nvidia Loading…
4 tasks
3
fix: fix github to myst-parser admonition conversion CI:docs Run doctest documentation Improvements or additions to documentation
#1224 opened Sep 29, 2025 by terrykong Loading…
11
Tk/slurm bisect documentation Improvements or additions to documentation
#1223 opened Sep 29, 2025 by terrykong • Draft
Set attention_mask to None by default. CI:L1 Run doctests, unit tests, and functional tests
#1213 opened Sep 26, 2025 by joyang-nv • Draft
4 tasks
5
draft feat: KV cache quantization support in fp8 rollout in GRPO Low Precision
#1212 opened Sep 26, 2025 by sharonyu-115 Loading…
4 tasks
1
7
feat: Multi-turn tool calling on BFCLv3 dataset community-request documentation Improvements or additions to documentation
#1207 opened Sep 25, 2025 by slikhite-1 Loading…
49
feat: Compute entropy across full vocab for logging r0.4.0
#1200 opened Sep 24, 2025 by parthchadha Loading…
4 tasks
6
chore: Bump vllm and ray
#1199 opened Sep 24, 2025 by guyueh1 Loading…
4 tasks
2
feat: [do not merge] Fp8 training kitchen documentation Improvements or additions to documentation
#1197 opened Sep 24, 2025 by guyueh1 • Draft
4 tasks
ci: Test runner CI:L1 Run doctests, unit tests, and functional tests CI Relating to CI
#1196 opened Sep 23, 2025 by chtruong814 Loading…
4 tasks
2
feat: FP8 rollout in GRPO for MoE models
#1175 opened Sep 21, 2025 by guyueh1 Loading…
4 tasks
1
@guyueh1
11
refactor: unify get_logprobs() and score() logic in dtensor CI:L1 Run doctests, unit tests, and functional tests
#1173 opened Sep 21, 2025 by RayenTian Loading…
1
@RayenTian
7
fix: simplified megatron to hf conversion script r0.4.0
#1169 opened Sep 20, 2025 by ahmadki Loading…
4 tasks
1
6
DSV3 feat branch
#1160 opened Sep 18, 2025 by joyang-nv • Draft
4 tasks
10
feat: Add Penguin env
#1156 opened Sep 18, 2025 by bxyu-nvidia Loading…
4 tasks
18
Update mcore / mbridge 0917
#1150 opened Sep 17, 2025 by yaoyu-33 Loading…
4 tasks
7
fix: Fix and add new tests for FLOPs accountant CI:L1 Run doctests, unit tests, and functional tests r0.4.0
#1149 opened Sep 17, 2025 by ybgao-nvidia Loading…
4 tasks
1
@ybgao-nvidia
10
ProTip! What’s not been updated in a month: updated:<2025-09-02.