[XPU] Add online fp8 quantization test (#44513)
Signed-off-by: Yan Ma <yan.ma@intel.com>
This commit is contained in:
@@ -10,15 +10,15 @@ from vllm.platforms import current_platform
|
||||
|
||||
|
||||
def is_quant_method_supported(quant_method: str) -> bool:
|
||||
# Currently, all quantization methods require Nvidia or AMD GPUs
|
||||
if not (current_platform.is_cuda() or current_platform.is_rocm()):
|
||||
# Currently, quantization tests only run GPUs
|
||||
if current_platform.is_cpu():
|
||||
return False
|
||||
|
||||
try:
|
||||
current_platform.verify_quantization(quant_method)
|
||||
except ValueError:
|
||||
return False
|
||||
|
||||
if current_platform.is_xpu():
|
||||
return True
|
||||
capability = current_platform.get_device_capability()
|
||||
assert capability is not None
|
||||
|
||||
|
||||
Reference in New Issue
Block a user