Skip to content

Pull requests: llm-d/llm-d-kv-cache

New pull request New
20 Open 395 Closed
20 Open 395 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

PVC Evictor: add crawler tests, CI, and docs after layout changes size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
#618 opened May 29, 2026 by guygir Collaborator Loading…
feat(kvevents): parse HMA KV event metadata size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
#612 opened May 26, 2026 by sagearc Collaborator Loading…
7
feat: batch KV block copies via cudaMemcpyBatchAsync in fs connector size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
#607 opened May 26, 2026 by kfirtoledo Collaborator Loading…
2 tasks done
5
deps(go): bump the go-dependencies group across 1 directory with 13 updates dependencies Pull requests that update a dependency file size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
#606 opened May 26, 2026 by dependabot Bot Loading…
1
feat: Add PVC evictor BlockRemoved events size/XL Denotes a PR that changes 500-999 lines, ignoring generated files.
#605 opened May 25, 2026 by albertoperdomo2 Contributor Loading…
8
fix(sglang): decode token_ids as typed []uint32 with bigram support size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
#603 opened May 23, 2026 by ryanx-sir Loading…
4 tasks done
6
Better tmp file handling (#2) size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
#600 opened May 20, 2026 by Prgrmman Loading…
3
fix: remove MIN_STAGING_BUFFER_SIZE that inflates KV cache on disk size/XS Denotes a PR that changes 0-9 lines, ignoring generated files.
#589 opened May 17, 2026 by Jwrede Loading…
1 of 2 tasks
1
Add SHA256-CBOR hashing algorithm for token processor with extra keys… size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files.
#587 opened May 15, 2026 by leipanhz Loading…
1
ci: Wire fs_backend Python tests into CI lgtm Looks good to me, indicates that a PR is ready to be merged. size/M Denotes a PR that changes 30-99 lines, ignoring generated files.
#578 opened May 7, 2026 by albertoperdomo2 Contributor Loading…
3
build: pin protoc for protobuf generation size/S Denotes a PR that changes 10-29 lines, ignoring generated files.
#576 opened May 6, 2026 by yankay Collaborator Loading…
6
fix: avoid LRU promotion on Lookup to unblock concurrent readers lifecycle/stale size/S Denotes a PR that changes 10-29 lines, ignoring generated files.
#562 opened May 1, 2026 by vMaroon Member • Draft
5
Add redis lookup and improve nixl lookup size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
#552 opened Apr 28, 2026 by effi-ofer Contributor Loading…
19
feat: Add RenderBatchCompletion RPC for multi-prompt tokenization size/XL Denotes a PR that changes 500-999 lines, ignoring generated files.
#538 opened Apr 24, 2026 by albertoperdomo2 Contributor Loading…
3
fix(e2e): fix container image build platform for macOS/non-Linux hosts lgtm Looks good to me, indicates that a PR is ready to be merged. size/S Denotes a PR that changes 10-29 lines, ignoring generated files.
#535 opened Apr 23, 2026 by gyliu513 Contributor Loading…
1
10
Add Hybrid Multi-head Attention (HMA) support for KV-Cache scoring lifecycle/rotten size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files.
#533 opened Apr 19, 2026 by kapiljain1989 Loading…
2
feat: Add HMA support to FS connector size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files.
#476 opened Mar 29, 2026 by kfirtoledo Collaborator Loading…
4 tasks done
1
17
deps(go): bump google.golang.org/grpc from 1.77.0 to 1.79.3 dependencies Pull requests that update a dependency file lifecycle/rotten
#438 opened Mar 19, 2026 by dependabot Bot Loading…
7
feat:add support to invalidate KV cache via AllBlocksCleared event size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files.
#437 opened Mar 18, 2026 by yash9263 Loading…
45
Add DP-aware routing support to KVEvents and indexing pipeline size/XL Denotes a PR that changes 500-999 lines, ignoring generated files.
#370 opened Feb 28, 2026 by satyamg1620 Loading…
39
ProTip! Follow long discussions with comments:>50.