enable ut test for xpu devices (#11712)

Co-authored-by: jundu <jun.du@intel.com>
Co-authored-by: Gao, Pengfei <pengfei.gao@intel.com>

test/registered/quant/test_awq_dequant.py

@@ -16,12 +16,13 @@ from sglang.srt.layers.quantization.awq_triton import (
     awq_dequantize_triton,
     awq_gemm_triton,
 )
+from sglang.srt.utils import get_device
 from sglang.test.ci.ci_register import register_amd_ci
 from sglang.test.test_utils import CustomTestCase
 
 register_amd_ci(est_time=2, suite="stage-a-test-1-amd")
 
-device = "cuda"
+device = get_device()
 
 
 def reverse_awq_order(t: torch.Tensor) -> torch.Tensor:

test/registered/quant/test_gptqmodel_dynamic.py

@@ -5,7 +5,7 @@ import requests
 import torch
 
 from sglang.srt.server_args import set_global_server_args_for_scheduler
-from sglang.srt.utils import kill_process_tree
+from sglang.srt.utils import get_device, kill_process_tree
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.test_utils import (
     DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
@@ -49,7 +49,7 @@ def check_quant_method(model_path: str, use_marlin_kernel: bool):
     model_config = ModelConfig.from_server_args(server_args)
 
     load_config = LoadConfig()
-    device_config = DeviceConfig("cuda")
+    device_config = DeviceConfig(get_device())
     model = get_model(
         model_config=model_config, load_config=load_config, device_config=device_config
     )