[Bugfix] Fix broken CPU compressed-tensors test (vllm-project#11338)

Signed-off-by: Isotr0py <[email protected]>
ruisearch42 · Dec 19, 2024 · 276738c · 276738c
1 parent cdf22af
commit 276738c
Showing 1 changed file with 2 additions and 4 deletions.
diff --git a/vllm/model_executor/layers/quantization/utils/w8a8_utils.py b/vllm/model_executor/layers/quantization/utils/w8a8_utils.py
@@ -11,8 +11,7 @@
 
 
 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()