From 276738ce0f6aac48ace36bc79aa4a0765fccdfb2 Mon Sep 17 00:00:00 2001 From: Isotr0py Date: Fri, 20 Dec 2024 01:37:31 +0800 Subject: [PATCH] [Bugfix] Fix broken CPU compressed-tensors test (#11338) Signed-off-by: Isotr0py <2037008807@qq.com> --- vllm/model_executor/layers/quantization/utils/w8a8_utils.py | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/vllm/model_executor/layers/quantization/utils/w8a8_utils.py b/vllm/model_executor/layers/quantization/utils/w8a8_utils.py index d77722499d0e9..d89071f30a549 100644 --- a/vllm/model_executor/layers/quantization/utils/w8a8_utils.py +++ b/vllm/model_executor/layers/quantization/utils/w8a8_utils.py @@ -11,8 +11,7 @@ TORCH_DEVICE_IDENTITY = torch.ones(1, dtype=torch.float32) def sparse_cutlass_supported() -> bool: - # sparse cutlass is not supported on Rocm - if current_platform.is_rocm(): + if not current_platform.is_cuda(): return False capability_tuple = current_platform.get_device_capability() @@ -22,8 +21,7 @@ def sparse_cutlass_supported() -> bool: def cutlass_fp8_supported() -> bool: - # cutlass is not supported on Rocm - if current_platform.is_rocm(): + if not current_platform.is_cuda(): return False capability_tuple = current_platform.get_device_capability()