Skip to content

Pull requests: RBLN-SW/vllm-rbln

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
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort

Pull requests list

fix(worker): unquote CompilationTimes return annotation
#673 opened Jun 12, 2026 by rebel-jinhwan Contributor Loading…
other(bump): bump vllm to 0.22.0 with optimum
#671 opened Jun 11, 2026 by rebel-eunji Collaborator Loading…
1 of 13 tasks
other(bump): make vllm-rbln compatible to vllm 0.22.0
#669 opened Jun 10, 2026 by rebel-jinhwan Contributor Draft
13 tasks
refactor(model): sync whisper with SupportsTranscription interface (WIP)
#668 opened Jun 10, 2026 by rebel-eunji Collaborator Loading…
13 tasks
refactor(multimodal): follow the multimodal interface of upstream vllm
#664 opened Jun 10, 2026 by rebel-eunji Collaborator Loading…
2 of 13 tasks
feature: select kernel compute dtype via RBLN_COMP_DTYPE torch.compile torch.compile based implementation
#654 opened Jun 8, 2026 by rebel-wonsubkim Contributor Loading…
1 of 13 tasks
fix: apply default batch size of 1 across all UsageContext cases
#652 opened Jun 8, 2026 by rebel-eunji Collaborator Loading…
13 tasks
refactor: reorganize RBLN runtime patching and model runner flow
#651 opened Jun 8, 2026 by junstar92 Collaborator Draft
35 of 69 tasks
feature: cross-block no-spec fallback for variable-length spec decoding proposers torch.compile torch.compile based implementation
#649 opened Jun 6, 2026 by rebel-wonsubkim Contributor Loading…
2 of 13 tasks
other(torch_compile): parametrize e2e test over RBLN_WEIGHT_FREE (wf_on xfail) torch.compile torch.compile based implementation
#647 opened Jun 5, 2026 by rebel-jinhwan Contributor Loading…
1 of 13 tasks
feature(warmup): offload host regions during device_tensor warm-up
#646 opened Jun 5, 2026 by rebel-jonghewk Collaborator Loading…
1 task done
fix: apply classifier activation in RBLNClassifierPooler
#644 opened Jun 4, 2026 by rebel-thkim Contributor Loading…
fix: pass hf_config to optimum compilation
#643 opened Jun 4, 2026 by rebel-thkim Contributor Loading…
feature(nixl): support nixl-rbln
#640 opened Jun 2, 2026 by rebel-yskim Contributor Draft
13 tasks
model: Gemma4 31B E2B and E4B torch.compile torch.compile based implementation
#632 opened May 29, 2026 by pei0033 Collaborator Loading…
4 of 13 tasks
feature: implement vllm benchmark tracing for analysis
#628 opened May 28, 2026 by rebel-wonsubkim Contributor Loading…
1 of 13 tasks
model: support deepseek v3
#624 opened May 26, 2026 by rebel-kblee Contributor Draft
1 of 13 tasks
[DO NOT MERGE] model: support gemma4
#622 opened May 26, 2026 by rebel-eunji Collaborator Draft
13 tasks
feature: introduce mega cache
#619 opened May 22, 2026 by rebel-jonghewk Collaborator Loading…
3 tasks done
other: switch vllm source to rebellions internal pypi
#614 opened May 19, 2026 by rebel-minhopark Loading…
3 of 13 tasks
other: bump to match upstream v0.19.1 (optimum)
#603 opened May 13, 2026 by rebel-eunji Collaborator Loading…
1 of 13 tasks
feat: enable request-reordering in optimum-compiled models
#596 opened May 11, 2026 by rebel-thkim Contributor Draft
13 tasks
other: improve optimum-compile
#591 opened May 7, 2026 by rebel-seinpark Collaborator Loading…
13 tasks
refactor: ci visibility improvement
#584 opened May 4, 2026 by rebel-seinpark Collaborator Draft
13 tasks
ProTip! Exclude everything labeled bug with -label:bug.