Skip to content

Pull requests: redai-infra/Relax

New pull request New
61 Open 79 Closed
61 Open 79 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(opd): [Task.39] EOPD cross-architecture distillation (arXiv:2603.07079)
#274 opened Aug 14, 2026 by buaacoder Loading…
1
[Task32] feat: fla mcore chunkwise cp stage 3
#273 opened Aug 13, 2026 by jambow0320 Loading…
3
【Task 40】Feat(p3o): add ESS-adaptive policy optimization
#272 opened Aug 13, 2026 by DreamEnding Contributor Loading…
3 of 5 tasks
1
3
fix(sglang): clamp deterministic sampler endpoints
#271 opened Aug 13, 2026 by DreamEnding Contributor Loading…
3 of 5 tasks
3
feat: add MemAgent multi-turn training example
#270 opened Aug 13, 2026 by TabsPhasers • Draft
9 of 16 tasks
feat(rollout): add SGLang native group sampling
#269 opened Aug 13, 2026 by overloadedHenry Contributor Loading…
9 tasks done
1
5
fix(lora): inline adapter tensors into the engine payload
#265 opened Aug 12, 2026 by SakaiXue6666 Loading…
4 of 11 tasks
fix(lora): write exported adapters in PEFT's key layout
#262 opened Aug 11, 2026 by SakaiXue6666 Loading…
4 of 11 tasks
fix(lora): expand path-pattern target modules to HF names
#261 opened Aug 11, 2026 by SakaiXue6666 Loading…
4 of 11 tasks
【Task.31】feat(sft): add reward modeling and preference evaluation
#258 opened Aug 11, 2026 by A-Words Contributor Loading…
6 of 11 tasks
【Task.31】feat(sft): add pair-aware DPO training
#257 opened Aug 11, 2026 by A-Words Contributor Loading…
6 of 11 tasks
feat(data-plane): enable RDMA transport for TransferQueue
#256 opened Aug 11, 2026 by overloadedHenry Contributor Loading…
14 of 16 tasks
1
8
Feat/task25 mixture of lora
#255 opened Aug 10, 2026 by pophirasawa Loading…
15 of 20 tasks
1
[Task32] feat: fla mcore chunkwise cp stage 2
#254 opened Aug 10, 2026 by jambow0320 Loading…
5
perf(megatron): skip zero-advantage backward
#253 opened Aug 8, 2026 by xiaoh1024 Contributor Loading…
4 of 5 tasks
1
5
[Task32] feat: fla mcore chunkwise cp stage 1
#251 opened Aug 7, 2026 by jambow0320 Loading…
2
【Task.33】feat: Add Reproducibility bundle
#248 opened Aug 5, 2026 by DreamEnding Contributor Loading…
6 of 11 tasks
7
【Task.01】多模态 Dense
#243 opened Aug 5, 2026 by A-Words Contributor Loading…
8 tasks done
1
1
perf: Carry-Aware Cross-Version KV Continuation + Hybrid DCS
#242 opened Aug 5, 2026 by jeason-djs Loading…
5 tasks done
1
15
【No.30】feat(dr-grpo): add Dr.GRPO support
#239 opened Aug 4, 2026 by idvzchusvol Loading…
5 of 11 tasks
17
feat(klx):add klx XPU training support for Qwen3-9B sft and Qwen3.6-2…
#238 opened Aug 4, 2026 by ChinaCYong Loading…
11 tasks
No.41 SDPO实现与复现
#237 opened Aug 4, 2026 by ZiyiTsang • Draft
6 of 8 tasks
3
perf(megatron): reuse process groups for colocate weight sync
#229 opened Aug 3, 2026 by ldemon2333 Loading…
3 of 12 tasks
3
feat(klx):add qwen35-9B-vl script
#227 opened Aug 3, 2026 by yeqinghe609 Contributor Loading…
11 tasks
【Task 22】perf(hybrid): 降低纯文本 Hybrid-async 权重发布开销
#211 opened Aug 2, 2026 by MaybeIcanShow Contributor Loading…
1
22
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.