[Bugfix] Refactor to support DP parallel in R3 (#32306)

Signed-off-by: xhx1022 <1737006628@qq.com>
Co-authored-by: arlenxu <arlenxu@tencent.com>
This commit is contained in:
Hongxin Xu
2026-01-16 15:13:58 +08:00
committed by GitHub
co-authored by arlenxu
parent 14ce524249
commit 03da3b52ef
3 changed files with 35 additions and 26 deletions
+1 -3
View File
@@ -5675,12 +5675,10 @@ class GPUModelRunner(
self.kv_cache_config.num_blocks // len(self.kv_cache_config.kv_cache_groups)
+ 1
) * block_size
routed_experts_capturer.init_buffer(
max_num_batched_tokens=self.scheduler_config.max_num_batched_tokens,
max_num_kv_tokens=self.max_num_kv_tokens,
model_config=self.model_config,
instance_id=self.vllm_config.instance_id,
vllm_config=self.vllm_config,
)
def may_add_encoder_only_layers_to_kv_cache_config(self) -> None: