Skip to content

Pull requests: alibaba/rtp-llm

New pull request New
159 Open 1,035 Closed
159 Open 1,035 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(ci-gate): explain fork rerun/approval denials instead of failing silently
#1324 opened Aug 23, 2026 by LLLLKKKK Collaborator Loading…
2 of 5 tasks
feat: enable Qwen3.5 W8A8 integration
#1323 opened Aug 20, 2026 by Tanmo-ai Contributor Loading…
132
ci: add post-merge K3 two-host smoke for feat/k3_dev
#1322 opened Aug 19, 2026 by lhcezx Collaborator • Draft
1
feat(qwen35): enable PPU runtime and safe CUDA graph replay
#1321 opened Aug 19, 2026 by Tanmo-ai Contributor Loading…
10
fix(cache): preserve aggregate MTP layout strides
#1320 opened Aug 19, 2026 by Honglei-Qiu Loading…
15
fix(indexer): keep the head gate off the fp32 activation expansion
#1319 opened Aug 19, 2026 by Honglei-Qiu Loading…
14
feat(quant): support compressed-tensors W8A8 INT8 checkpoints
#1317 opened Aug 19, 2026 by Tanmo-ai Contributor Loading…
201
Feature/qwen35 mtp pd cg smoke
#1316 opened Aug 19, 2026 by siluzhou Collaborator Loading…
78
feat(trace): end-to-end OpenTelemetry request tracing (combined review PR)
#1315 opened Aug 19, 2026 by SAzwj Contributor Loading…
157
fix(indexer): route MTP target verify through the paged TopK
#1314 opened Aug 18, 2026 by Honglei-Qiu Loading…
50
feat: add TorchSpec target backend support
#1312 opened Aug 18, 2026 by devillove084 Loading…
274
fix(speculative): report greedy proposal distribution
#1311 opened Aug 18, 2026 by Honglei-Qiu Loading…
24
feat(trace): add the C++ OpenTelemetry runtime and stream progress milestones
#1310 opened Aug 18, 2026 by SAzwj Contributor Loading…
114
feat(frontend): report request and cache metrics
#1309 opened Aug 17, 2026 by Nancheng-11 Collaborator Loading…
142
fix(moe): reject normal DeepEP CUDA Graph
#1308 opened Aug 17, 2026 by Honglei-Qiu Loading…
23
feat(glm5): Optimize prefill TopK postprocessing, Q-RoPE writes, CP restore, and routing
#1307 opened Aug 17, 2026 by MTaoYue Loading…
45
fix: tokenize request stop words in context; fix frontend_app_test port conflicts
#1306 opened Aug 17, 2026 by SAzwj Contributor Loading…
49
Feature/glm47 fix decode graph buffer capacity
#1304 opened Aug 17, 2026 by fxp19950422 Loading…
92
[Feat][Perf] DeepSeek-V4-Flash on Hopper: FP8 routed experts, plus a decode/prefill optimization pass
#1303 opened Aug 15, 2026 by ntguojiarui Loading…
6 of 8 tasks
104
feat: add vllm fa4 impl for spec decode target verify and draft model prefill forward
#1301 opened Aug 14, 2026 by HongminTan Collaborator Loading…
236
fix(qwen3.5): stabilize prefill context parallelism
#1299 opened Aug 14, 2026 by zhyuan020 Loading…
133
refactor(cache): converge independent group pools and harden grouped cache contracts
#1298 opened Aug 14, 2026 by Adrenaline-S Collaborator Loading…
56
feat(runtime): add deferred out-of-tree backend registration
#1297 opened Aug 13, 2026 by Tanmo-ai Contributor Loading…
58
fix: correct MTP speculative sampling semantics
#1295 opened Aug 13, 2026 by Vinkle-hzt Collaborator Loading…
11 tasks done
119
feat: add forward warmup for PD engines
#1294 opened Aug 13, 2026 by ydshi0 Collaborator Loading…
216
ProTip! Exclude everything labeled bug with -label:bug.