add xpu support HFLM (#3211)

Signed-off-by: Liu, Kaixuan <kaixuan.liu@intel.com>

lm_eval/models/huggingface.py

@@ -124,14 +124,22 @@ class HFLM(TemplateLM):
             assert isinstance(pretrained, str)
             assert isinstance(batch_size, (int, str))
 
-            gpus = torch.cuda.device_count()
             accelerator_kwargs = InitProcessGroupKwargs(timeout=timedelta(weeks=52))
             accelerator = Accelerator(kwargs_handlers=[accelerator_kwargs])
             if accelerator.num_processes > 1:
                 self.accelerator = accelerator
 
-            if "npu" in accelerator.device.type:
+            # Detect device count based on accelerator device type
+            device_type = accelerator.device.type
+            if "cuda" in device_type:
+                gpus = torch.cuda.device_count()
+            elif "npu" in device_type:
                 gpus = torch.npu.device_count()
+            elif "xpu" in device_type:
+                gpus = torch.xpu.device_count()
+            else:
+                # Fallback to CUDA count for compatibility
+                gpus = torch.cuda.device_count()
 
             # using one process with no model parallelism
             if not (parallelize or accelerator.num_processes > 1):
@@ -141,6 +149,7 @@ class HFLM(TemplateLM):
                     + [f"cuda:{i}" for i in range(gpus)]
                     + ["mps", "mps:0"]
                     + [f"npu:{i}" for i in range(gpus)]
+                    + [f"xpu:{i}" for i in range(gpus)]
                 )
                 if device and device in device_list:
                     self._device = torch.device(device)