fix xpu device_id AttributeError issue (#7488)

# Reproduce
w/ PyTorch 2.8
```
$ git clone https://github.com/huggingface/trl.git
$ cd ./trl
$ accelerate launch     --config_file examples/accelerate_configs/deepspeed_zero3.yaml     examples/scripts/sft_gpt_oss.py     --torch_dtype bfloat16     --model_name_or_path openai/gpt-oss-20b     --packing true packing_strategy wrapped     --run_name 20b-full-eager     --attn_implementation sdpa     --dataset_num_proc 6     --dataset_name HuggingFaceH4/Multilingual-Thinking     --gradient_checkpointing     --max_length 4096     --per_device_train_batch_size 1     --num_train_epochs 1     --logging_steps 1     --warmup_ratio 0.03     --lr_scheduler_type cosine_with_min_lr     --lr_scheduler_kwargs '{"min_lr_rate": 0.1}'     --output_dir gpt-oss-20b-multilingual-reasoner     --report_to trackio     --seed 42
```

# Issue

> File "/workspace/accelerate/src/accelerate/state.py", line 216, in
__init__
> dist.init_distributed(dist_backend=self.backend,
auto_mpi_discovery=False, **kwargs)
> File "/usr/local/lib/python3.12/dist-packages/deepspeed/comm/comm.py",
line 854, in init_distributed
> cdb = TorchBackend(dist_backend, timeout, init_method, rank,
world_size)
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> File
"/usr/local/lib/python3.12/dist-packages/deepspeed/comm/torch.py", line
120, in __init__
> self.init_process_group(backend, timeout, init_method, rank,
world_size)
> File
"/usr/local/lib/python3.12/dist-packages/deepspeed/comm/torch.py", line
164, in init_process_group
>     torch.distributed.init_process_group(backend, **kwargs)
> File
"/usr/local/lib/python3.12/dist-packages/torch/distributed/c10d_logger.py",
line 81, in wrapper
>     return func(*args, **kwargs)
>            ^^^^^^^^^^^^^^^^^^^^^
> File
"/usr/local/lib/python3.12/dist-packages/torch/distributed/c10d_logger.py",
line 95, in wrapper
>     func_return = func(*args, **kwargs)
>                   ^^^^^^^^^^^^^^^^^^^^^
> File
"/usr/local/lib/python3.12/dist-packages/torch/distributed/distributed_c10d.py",
line 1685, in init_process_group
>     if device_id is not None and device_id.type != "cpu":
> AttributeError: 'device' object has no attribute 'type'

# Root Cause
`torch.xpu.device` in PyTorch is a context manager in PyTorch rather
than a device class, it doesn't have attribute `type`

# Fix
switch to use `torch.device`

Signed-off-by: Yao, Matrix <matrix.yao@intel.com>
Co-authored-by: Olatunji Ruwase <tunji.ruwase@snowflake.com>

accelerator/xpu_accelerator.py

@@ -61,7 +61,7 @@ class XPU_Accelerator(DeepSpeedAccelerator):
         return 'xpu:{}'.format(device_index)
 
     def device(self, device_index=None):
-        return torch.xpu.device(device_index)
+        return torch.device('xpu', device_index)
 
     def set_device(self, device_index):
         torch.xpu.set_device(device_index)