AlexHuang 0fec3d652b [Bugfix][KV Offload] Centralize shared mmap cleanup in CPU worker (#51622) ...

Signed-off-by: Alex <jihui.huang@daocloud.io>

2026-08-11 13:22:15 +03:00

..

attention

[Kimi-K3] DCP support (#50484)

2026-08-10 12:54:26 -07:00

core

[Bugfix][Core] Preserve Mamba running CoW after external hits (#51766)

2026-08-11 18:15:24 +08:00

cudagraph

…

determinism

…

distributed

…

e2e

[CI] Solidify speculative decoding E2E coverage (#50713)

2026-08-10 19:56:39 -06:00

ec_connector

…

engine

…

executor

…

fault_tolerance

…

kv_connector

[KV Connector][Offloading] Keep per-layer KV registration when canonical_layout is requested (#51688)

2026-08-11 06:03:59 +03:00

kv_offload

[Bugfix][KV Offload] Centralize shared mmap cleanup in CPU worker (#51622)

2026-08-11 13:22:15 +03:00

logits_processors

[Bugfix] Preserve non-logitproc entry points in tests (#51097)

2026-08-10 18:14:53 -04:00

metrics

…

sample

…

shutdown

…

simple_kv_offload

…

spec_decode

[Model Runner V2] Fix -1 placeholder draft token ids in rejection sam… (#50939)

2026-08-06 16:32:00 -07:00

streaming_input

…

structured_output

[Bugfix][Structured Output] Mask request stop tokens in xgrammar until grammar terminates (#49227)

2026-08-10 14:29:20 -04:00

tracing

…

worker

[MRV2][Spec] Fuse AR speculator multi-step decodes back into one CUDA graph (#46849)

2026-08-11 15:06:22 +08:00

__init__.py

…

test_kv_cache_spec_registry.py

…

test_oracle.py

…

test_outputs.py

…

test_request.py

…

test_serial_utils.py

…

test_tensor_ipc_queue.py

…

utils.py

…