From 9da51c77a9707126c82f5db424b12233c9dbc093 Mon Sep 17 00:00:00 2001 From: David-Wen <18927700430@163.com> Date: Sat, 20 Sep 2025 04:44:24 +0800 Subject: [PATCH] Fix: Correct FusedMoE layer reference in auto_round quantization (#24818) Signed-off-by: David-Wen <18927700430@163.com> Signed-off-by: Michael Goin Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com> Co-authored-by: Michael Goin Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Signed-off-by: yewentao256 --- vllm/model_executor/layers/quantization/auto_round.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/vllm/model_executor/layers/quantization/auto_round.py b/vllm/model_executor/layers/quantization/auto_round.py index 1ca92273430dd..bf5141fa48943 100644 --- a/vllm/model_executor/layers/quantization/auto_round.py +++ b/vllm/model_executor/layers/quantization/auto_round.py @@ -241,7 +241,7 @@ class AutoRoundConfig(QuantizationConfig): if isinstance(layer, FusedMoE): if use_marlin: - return AWQMoEMethod(quant_args_marlin, layer.moe) + return AWQMoEMethod(quant_args_marlin, layer.moe_config) from vllm.model_executor.layers.quantization.moe_wna16 import ( MoeWNA16Config) @@ -327,7 +327,7 @@ class AutoRoundConfig(QuantizationConfig): if isinstance(layer, FusedMoE): if use_marlin: - return GPTQMarlinMoEMethod(quant_args_marlin, layer.moe) + return GPTQMarlinMoEMethod(quant_args_marlin, layer.moe_config) else: from vllm.model_executor.layers.quantization.moe_wna16 import ( MoeWNA16Config)