Skip to content

Pull requests: alibaba/rtp-llm

New pull request New
177 Open 1,056 Closed
177 Open 1,056 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: overlap CUDA graph replay preparation on PRO5000
#1365 opened Sep 1, 2026 by siluzhou Collaborator Loading…
1
perf(sm120): use Triton FP8 MoE for small decode batches
#1364 opened Sep 1, 2026 by siluzhou Collaborator Loading…
11
feat(sm120): 将 PR #1223 的 FlashInfer B12X NVFP4 MoE 支持移植到 main
#1363 opened Aug 31, 2026 by ykykzq Loading…
58
Feat/sm120 deepgemm fp8
#1362 opened Aug 31, 2026 by parkerpang Collaborator Loading…
57
feat: remove redundant fusedQKV bias transpose kernel in vision bert …
#1361 opened Aug 31, 2026 by junna2016 Collaborator Loading…
11
feat: support full prefill CUDA graphDevelop/full prefill cuda graph
#1360 opened Aug 31, 2026 by siluzhou Collaborator Loading…
40
feat(flexlb): support Nacos local standby hot updates
#1358 opened Aug 31, 2026 by Martin7-1 Collaborator Loading…
25
feat(flexlb): configure vllm block hashing
#1357 opened Aug 31, 2026 by Martin7-1 Collaborator Loading…
42
feat(rocm): add AITER FlyDSL GDN decode adapter
#1355 opened Aug 31, 2026 by junna2016 Collaborator Loading…
67
feat(loader): integrate bounded rank-local FastSafeTensors copyout
#1354 opened Aug 30, 2026 by ABNER-1 Collaborator Loading…
216
feat: add RDMA transport for ViT output
#1353 opened Aug 30, 2026 by ydshi0 Collaborator Loading…
197
mimo v2.5 support
#1351 opened Aug 29, 2026 by RK-PHG Loading…
31
feat(sm120): enable DSv4 Flash 0731 inference support
#1350 opened Aug 28, 2026 by MMadhatter Collaborator Loading…
154
feat(ascend): add Huawei Ascend NPU support
#1349 opened Aug 28, 2026 by ningweikang Loading…
83
Dev/main block tree cache 0828
#1348 opened Aug 28, 2026 by MMeecatfish Collaborator Loading…
31
feat: add cross-node CPU TP broadcast
#1347 opened Aug 28, 2026 by xinfei-shi Collaborator Loading…
20
feat: monitor repeated model output
#1346 opened Aug 28, 2026 by xinfei-shi Collaborator Loading…
21
feat(rocm): add shape-based AITER FMoE tuning
#1345 opened Aug 28, 2026 by yzyDavid Collaborator Loading…
70
fix: improve RPC/cache-store diagnostics and xgrammar loading
#1344 opened Aug 28, 2026 by xinfei-shi Collaborator Loading…
53
feat: qwen35 share expert and deepep combine overlap
#1343 opened Aug 28, 2026 by HongminTan Collaborator Loading…
41
feat(cache): add native KVCM cache event publisher
#1340 opened Aug 27, 2026 by putaopi7 Collaborator Loading…
81
fix: keep decode rope positions on CUDA
#1337 opened Aug 26, 2026 by siluzhou Collaborator Loading…
123
feat: add KVCM fallback to master client
#1335 opened Aug 26, 2026 by Zhanghyi Loading…
123
Drop deprecated DeepSeek smoke cases, fix DSv4 CP prefill + thinking_mode
#1333 opened Aug 26, 2026 by netaddi Collaborator Loading…
4 of 5 tasks
213
feat(flexlb): support logical multi-engine workers
#1332 opened Aug 25, 2026 by Martin7-1 Collaborator Loading…
24
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.