Skip to content

Pull requests: alibaba/rtp-llm

New pull request New
126 Open 927 Closed
126 Open 927 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: restore cross-sequence ban for n returns
#1177 opened Jul 10, 2026 by Xavier2024P Contributor Loading…
1
Features/amd qwen
#1175 opened Jul 9, 2026 by siluzhou Collaborator Loading…
4
feat: support chunked prefill
#1174 opened Jul 8, 2026 by Jonathan-hwx Collaborator Loading…
12
feat(attn): dynamic decode attention backend selection at CUDA graph capture
#1173 opened Jul 8, 2026 by SAzwj Contributor Loading…
9
Feature/opt fifo scheduler main
#1171 opened Jul 8, 2026 by HoniiTro19 Contributor Loading…
10
feat(xpu): add Intel GPU (XPU) support
#1167 opened Jul 7, 2026 by aslanxie Loading…
17
Feat/adapt sm120 build
#1166 opened Jul 7, 2026 by parkerpang Collaborator Loading…
7
Feat/dsv4 on dev refactor block tree
#1165 opened Jul 7, 2026 by xiqi-lq Collaborator Loading…
47
Add newloader core framework
#1164 opened Jul 7, 2026 by Oneydauh Loading…
42
fix: sync mtp position ids with spec tokens
#1163 opened Jul 7, 2026 by HoniiTro19 Contributor Loading…
18
feat: add topology-aware sparse MLA candidate policy
#1162 opened Jul 7, 2026 by teerthsharma • Draft
23
bench: add topology KV candidate schedule benchmark
#1160 opened Jul 6, 2026 by teerthsharma Loading…
87
fix: avoid num_return_sequences prefill batch fanout
#1158 opened Jul 6, 2026 by Vinkle-hzt Collaborator Loading…
1
fix: read rope_theta/dtype from transformers>=5.2.0 nested config
#1156 opened Jul 1, 2026 by Jonathan-hwx Collaborator Loading…
3
Add KVS cache connector for KV cache reuse
#1154 opened Jul 1, 2026 by RulinJuice Collaborator Loading…
6
[WIP]add coordinated sleep/wake lifecycle for GPU resource parking
#1151 opened Jun 30, 2026 by MMadhatter Collaborator • Draft
7
fix: pass hw_kernel_config properly throughout py models
#1149 opened Jun 30, 2026 by zhangjianning-zjn Collaborator Loading…
1
feat: add Qwen3-VL custom ViT implementation for vision acceleration
#1144 opened Jun 29, 2026 by theNiemand Collaborator Loading…
4
Feat: Let RTP ROCm side use ATOM plugin interfaces
#1141 opened Jun 25, 2026 by zhaoan12-prc Collaborator Loading…
25
kvcache framework restructure
#1139 opened Jun 25, 2026 by Adrenaline-S Collaborator Loading…
30
feat(flexlb): batch scheduling infrastructure, strategy refactor, and error model hardening
#1138 opened Jun 24, 2026 by wzy-99 Collaborator Loading…
25
fix(bert): propagate multimodal features to Python BertModel forward
#1137 opened Jun 23, 2026 by parkerpang Collaborator Loading…
19
Feature/kvcm online optmizer
#1135 opened Jun 23, 2026 by YoungRX Collaborator Loading…
1
fix: enable model analyzer to resolve DeepSeek-V4 and Qwen3.5 models
#1130 opened Jun 22, 2026 by kwohting Loading…
3
feat(deps): pip dep unification + cu126/dead-config cleanup + WORKSPA…
#1124 opened Jun 22, 2026 by baohengyi Loading…
45
ProTip! Mix and match filters to narrow down what you’re looking for.