Skip to content

Pull requests: local-inference-lab/vllm

New pull request New
23 Open 156 Closed
23 Open 156 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

[GG] feat(dcp): support partial replicated sparse-indexer topology
#179 opened Jul 25, 2026 by voipmonitor Loading…
7
[GG] perf(dcp): merge sparse prefill top-k by row owner
#178 opened Jul 25, 2026 by voipmonitor Loading…
6
[GG] perf(dcp): preallocate budgeted CKV layer prefetch
#177 opened Jul 25, 2026 by voipmonitor Loading…
15
[GG] perf(indexer): shard sparse prefill queries and halve result traffic
#175 opened Jul 24, 2026 by voipmonitor Loading…
2
[GG] Fix MRV2 high-utilization startup memory profiling
#172 opened Jul 23, 2026 by voipmonitor Loading…
2
fix(attention): qualify compact NVFP4 MTP verifier route
#171 opened Jul 23, 2026 by yatesdr • Draft
2
fix(mrv2): reuse profiled CUDA graph pool
#168 opened Jul 22, 2026 by yatesdr Loading…
3
[GG] feat(kv-offload): bound filesystem tier capacity
#165 opened Jul 22, 2026 by yatesdr Loading…
3
[EXL3] Route prefill batches through the planned Trellis MoE (+58-64% prefill, stacked on #139)
#163 opened Jul 22, 2026 by davidsyoung Loading…
8
feat(dcp): add selected-record sparse CKV decode
#161 opened Jul 21, 2026 by FujitsuPolycom Loading…
4
feat(dcp): add native depth-N CKV prefetch
#160 opened Jul 21, 2026 by FujitsuPolycom Loading…
6
feat(kv-cache): support replicated MLA indexer with sharded CKV
#159 opened Jul 21, 2026 by FujitsuPolycom Loading…
5
[Core] Allow verified copy-based KV connectors with expandable segments
#158 opened Jul 21, 2026 by chriswritescode-dev Loading…
2
[GG] MLA: release absorbed B12X MXFP8 projection weights
#154 opened Jul 21, 2026 by voipmonitor Loading…
4
glm52: calibrated NVFP4 MLA KV outer scales + serve wiring
#145 opened Jul 20, 2026 by MadeBy561 Loading…
6
[Quant] Add EXL3 Trellis backend for rank-sliced GLM/DeepSeek MoE
#139 opened Jul 20, 2026 by brandonmmusic-max Loading…
26
spec_decode: consolidate DSpark correctness and load-aware serving
#109 opened Jul 17, 2026 by voipmonitor Loading…
11
[KV Offload] Align SWA store sparsity with local-cache retention semantics
#89 opened Jul 12, 2026 by procr1337 Loading…
2
MiniMax M3: fp8 (e4m3) index-K side cache (+61% long-context decode)
#75 opened Jul 5, 2026 by jarrelscy Loading…
4
spec_decode: propagate DCP size to V1 draft config (fixes MTP collapse under DCP)
#72 opened Jul 3, 2026 by m9e • Draft
4
fix: bypass reasoning gate for completions requests
#59 opened Jun 27, 2026 by JMPSequeira Loading…
2
fix: force xgrammar backend in from_sampling_params
#53 opened Jun 26, 2026 by JMPSequeira Loading…
10
[parser] Fix Qwen3 prompt-end reasoning check for multi-turn / tools
#46 opened Jun 25, 2026 by cyysky Loading…
4
ProTip! Add no:assignee to see everything that’s not assigned.