fems14 17dd9ae42c [0.11.0][bugfix]look up multi_tp key (#3699) (#3723) ...

### What this PR does / why we need it?
In multi-Tensor Parallel (TP) scenarios, the KV pool only queries the
first GPU card. When keys on other cards are released, the query result
still returns as successful, introducing accuracy issues. This PR
modifies the KV pool's query logic to check all cards, resolving this
problem.
### Does this PR introduce _any_ user-facing change?

### How was this patch tested?

- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

Signed-off-by: fems14 <1804143737@qq.com>

2025-10-24 18:22:45 +08:00

..

cpu_offload_manager

[Feature]cpu offload connector (#1659)

2025-09-23 14:25:05 +08:00

device_communicators

[MISC] Clean up torch_npu (#688)

2025-04-29 18:03:38 +08:00

mooncake

[0.11.0][bugfix]look up multi_tp key (#3699) (#3723)

2025-10-24 18:22:45 +08:00

__init__.py

【bugfix】fix connector register failed (#3335)

2025-10-09 21:09:54 +08:00

communicator.py

[2/N][Feat] Add MC2 communication method for MoE layers (#2469)

2025-08-26 19:05:23 +08:00

cpu_offload_connector.py

[KVCache] Refactor KVCache as page_size_bytes is ineffective (#3438)

2025-10-14 21:28:41 +08:00

llmdatadist_c_mgr_connector.py

[Refactor] Adapt deepseek-v3.2 to vllm 0.11.0 (#3432)

2025-10-15 17:48:58 +08:00

mooncake_connector.py

[0.11.0][Bugfix] fix delay free prefill req & D node support prefix cache (#3609)

2025-10-23 20:39:35 +08:00

mooncake_layerwise_connector.py

[Bugfix] fix ZeroDivisionError when prefill_tp_size > num_kv_head and fix tp_resharding README (#3437)

2025-10-15 08:45:44 +08:00

parallel_state.py

[Bugfix] TP size larger than KV cache head causes accuracy issues (#3366)

2025-10-11 11:22:23 +08:00

utils.py

KVCache Transfer via Layer-wise Strategy in Disaggregation (#2602)

2025-09-30 15:10:29 +08:00