Karan BansalandInokinoki 821fde2df4 [Bugfix] Fix xgrammar dtype mismatch on macOS CPU inference (#32384) ...

Signed-off-by: Karan Bansal <karanb192@gmail.com>
Co-authored-by: Inokinoki <inoki@inoki.cc>

2026-03-14 17:29:06 +00:00

..

attention

[Feat][Bugfix] Enable additional dimension for Flashinfer MLA and fix routing dtype (#36931)

2026-03-13 16:41:16 -07:00

core

fix(kv-cache): increase hybrid attention grouping threshold from 1.25 to 1.5 (#36684)

2026-03-12 23:28:27 -04:00

engine

[BugFix] Fix "DP Coordinator receives unexpected..." messages (#37008)

2026-03-13 23:18:05 +00:00

executor

[Bugfix] Surface exceptions from non-blocking execute_model in UniProcExecutor to avoid DP deadlocks (#35194)

2026-03-11 03:22:21 +00:00

kv_offload

[kv_offload+HMA][1/N]: Support multiple KV groups in OffloadingSpec (#36610)

2026-03-13 08:04:21 +00:00

metrics

[Metrics] Add Prometheus counters for Model FLOPs Utilization (MFU) (#30950)

2026-02-23 15:01:07 +00:00

pool

[Perf] Optimize compute maxsim using batched version, 3.2% E2E throughput improvement (#36710)

2026-03-12 08:37:01 +08:00

sample

[CI] Bump mypy version (#34950)

2026-03-04 20:55:11 +00:00

spec_decode

Kimi k2.5 MLA based eagle3 (#36361)

2026-03-11 11:36:11 -04:00

structured_output

[Bugfix] Fix xgrammar dtype mismatch on macOS CPU inference (#32384)

2026-03-14 17:29:06 +00:00

worker

[Model Runner V2] Add Support for XD-RoPE (#36817)

2026-03-14 09:27:55 -07:00

__init__.py

[V1] AsyncLLM Implementation (#9826)

2024-11-11 23:05:38 +00:00

cudagraph_dispatcher.py

[UX][Startup] Account for CUDA graphs during memory profiling (#30515)

2026-03-07 13:49:23 -08:00

kv_cache_interface.py

[BUGFIX][Mamba][Qwen3.5] Zero freed SSM cache blocks on GPU (#35219)

2026-03-10 03:32:20 -07:00

outputs.py

[KVConnector] Support worker -> scheduler metadata (#31964)

2026-03-11 17:36:37 +00:00

request.py

[Deprecation][1/2] Remove items deprecated in v0.18 (#36470)

2026-03-09 03:43:23 -07:00

serial_utils.py

[Bug] Fix Failure in /v1/chat/completions/render for Multimodal Requests (https://github.com/vllm-project/vllm/issues/35665) (#35684)

2026-03-14 14:10:11 +00:00

utils.py

[Frontend][Core] Re-add shutdown timeout - allowing in-flight requests to finish (#36666)

2026-03-13 12:10:06 -07:00