From 58a37ce1891e0e671191ba1fc9f1b6cbf60dab8d Mon Sep 17 00:00:00 2001 From: zouyida2052 Date: Sun, 19 Oct 2025 17:06:05 +0800 Subject: [PATCH] bugfix for mooncake (#3535) ### What this PR does / why we need it? bugfix for mooncake, remove useless judgement. ### How was this patch tested? by ci - vLLM version: v0.11.0rc3 - vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0 Signed-off-by: zouyida2052 --- vllm_ascend/distributed/mooncake_connector.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/vllm_ascend/distributed/mooncake_connector.py b/vllm_ascend/distributed/mooncake_connector.py index ebab077..15dee25 100644 --- a/vllm_ascend/distributed/mooncake_connector.py +++ b/vllm_ascend/distributed/mooncake_connector.py @@ -943,7 +943,7 @@ class MooncakeConnectorWorker: # kv_transfer variables self.vllm_config = vllm_config self.block_size = vllm_config.cache_config.block_size - if self.vllm_config.model_config.is_deepseek_mla or self.use_sparse: + if self.vllm_config.model_config.is_deepseek_mla: self.num_need_pulls = 1 else: num_d_block_heads = max(1,