From aaa0d5558707e73c4df98deac2491756725c9699 Mon Sep 17 00:00:00 2001 From: vllmellm Date: Tue, 4 Nov 2025 14:40:30 +0000 Subject: [PATCH] format Signed-off-by: vllmellm --- .../layers/quantization/kernels/scaled_mm/pytorch.py | 2 -- 1 file changed, 2 deletions(-) diff --git a/vllm/model_executor/layers/quantization/kernels/scaled_mm/pytorch.py b/vllm/model_executor/layers/quantization/kernels/scaled_mm/pytorch.py index 40e55cc97392c..10293c445a347 100644 --- a/vllm/model_executor/layers/quantization/kernels/scaled_mm/pytorch.py +++ b/vllm/model_executor/layers/quantization/kernels/scaled_mm/pytorch.py @@ -230,14 +230,12 @@ class RowWiseTorchScaledMMLinearKernel(TorchScaledMMLinearKernel): class ChannelWiseTorchScaledMMLinearKernel(TorchScaledMMLinearKernel): - @classmethod def can_implement(cls, c: FP8ScaledMMLinearLayerConfig) -> tuple[bool, str | None]: per_tensor_activation_scales = ( c.activation_quant_key.scale.group_shape.is_per_tensor() ) per_tensor_weight_scales = c.weight_quant_key.scale.group_shape.is_per_tensor() - if per_tensor_activation_scales and per_tensor_weight_scales: return ( False,