From fb22be5817cc772cd8bda02d73ca26bcac12751c Mon Sep 17 00:00:00 2001 From: Mengqing Cao Date: Thu, 27 Mar 2025 12:50:29 +0800 Subject: [PATCH] [moe][quant] add weight name case for offset (#15515) Signed-off-by: Mengqing Cao --- vllm/model_executor/layers/fused_moe/layer.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/vllm/model_executor/layers/fused_moe/layer.py b/vllm/model_executor/layers/fused_moe/layer.py index b72f51aa52bfa..711bdfd688501 100644 --- a/vllm/model_executor/layers/fused_moe/layer.py +++ b/vllm/model_executor/layers/fused_moe/layer.py @@ -699,8 +699,9 @@ class FusedMoE(torch.nn.Module): tp_rank=self.tp_rank) return - # Case weight scales and zero_points - if ("scale" in weight_name or "zero" in weight_name): + # Case weight scales, zero_points and offset + if ("scale" in weight_name or "zero" in weight_name + or "offset" in weight_name): # load the weight scales and zp based on the quantization scheme # supported weight scales/zp can be found in # FusedMoeWeightScaleSupported