From d19264ea95140763f77ebf59931ee8f63f7f41bb Mon Sep 17 00:00:00 2001 From: Isotr0py Date: Tue, 27 Jan 2026 11:22:55 +0800 Subject: [PATCH] fuse moe contiguous Signed-off-by: Isotr0py --- vllm/model_executor/layers/fused_moe/layer.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/vllm/model_executor/layers/fused_moe/layer.py b/vllm/model_executor/layers/fused_moe/layer.py index 38216350655..e3827500484 100644 --- a/vllm/model_executor/layers/fused_moe/layer.py +++ b/vllm/model_executor/layers/fused_moe/layer.py @@ -1140,7 +1140,7 @@ class FusedMoE(CustomOp): "CompressedTensorsWNA16MarlinMoEMethod", "CompressedTensorsWNA16MoEMethod", ): - loaded_weight = loaded_weight.t() + loaded_weight = loaded_weight.t().contiguous() if shard_id not in ("w1", "w2", "w3"): raise ValueError(f"shard_id must be ['w1','w2','w3'] but got {shard_id}.")