AttributeError: '_OpNamespace' '_C' object has no attribute 'get_cuda_view_from_cpu_tensor' when using vllm serve --cpu-offload-gb #869
Description
Describe the bug
Command: vllm serve Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound --max-model-len 3K --download-dir /media/ninemeow/Kioxia --dtype=half --enforce-eager --swap-space 5 --cpu-offload-gb 16
Log:
[W1017 21:10:16.915307772 OperatorEntry.cpp:218] Warning: Warning only once for all operators, other operators may also be overridden.
Overriding a previously registered kernel for the same operator and the same dispatch key
operator: aten::geometric_(Tensor(a!) self, float p, *, Generator? generator=None) -> Tensor(a!)
registered at /pytorch/build/aten/src/ATen/RegisterSchema.cpp:6
dispatch key: XPU
previous kernel: registered at /pytorch/aten/src/ATen/VmapModeRegistrations.cpp:37
new kernel: registered at /build/intel-pytorch-extension/build/Release/csrc/gpu/csrc/gpu/xpu/ATen/RegisterXPU_0.cpp:172 (function operator())
INFO 10-17 21:10:17 [init.py:225] Automatically detected platform xpu.
(APIServer pid=782393) INFO 10-17 21:10:19 [api_server.py:1876] vLLM API server version 0.11.1rc2.dev115+g3aeb19a39
(APIServer pid=782393) INFO 10-17 21:10:19 [utils.py:243] non-default args: {'model_tag': 'Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound', 'model': 'Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound', 'dtype': 'half', 'max_model_len': 3072, 'enforce_eager': True, 'download_dir': '/media/ninemeow/Kioxia', 'swap_space': 5.0, 'cpu_offload_gb': 16.0}
(APIServer pid=782393) INFO 10-17 21:10:21 [model.py:658] Resolved architecture: Qwen3NextForCausalLM
(APIServer pid=782393) WARNING 10-17 21:10:21 [logger.py:68] Casting torch.bfloat16 to torch.float16.
(APIServer pid=782393) INFO 10-17 21:10:21 [model.py:1745] Using max model len 3072
(APIServer pid=782393) INFO 10-17 21:10:22 [scheduler.py:225] Chunked prefill is enabled with max_num_batched_tokens=2048.
(APIServer pid=782393) INFO 10-17 21:10:22 [config.py:323] Disabling cascade attention since it is not supported for hybrid models.
(APIServer pid=782393) INFO 10-17 21:10:22 [config.py:439] Setting attention block size to 544 tokens to ensure that attention page size is >= mamba page size.
(APIServer pid=782393) INFO 10-17 21:10:22 [config.py:463] Padding mamba page size by 1.49% to ensure that mamba page size and attention page size are exactly equal.
[W1017 21:10:25.624681163 OperatorEntry.cpp:218] Warning: Warning only once for all operators, other operators may also be overridden.
Overriding a previously registered kernel for the same operator and the same dispatch key
operator: aten::geometric(Tensor(a!) self, float p, *, Generator? generator=None) -> Tensor(a!)
registered at /pytorch/build/aten/src/ATen/RegisterSchema.cpp:6
dispatch key: XPU
previous kernel: registered at /pytorch/aten/src/ATen/VmapModeRegistrations.cpp:37
new kernel: registered at /build/intel-pytorch-extension/build/Release/csrc/gpu/csrc/gpu/xpu/ATen/RegisterXPU_0.cpp:172 (function operator())
INFO 10-17 21:10:26 [init.py:225] Automatically detected platform xpu.
(EngineCore_DP0 pid=782659) INFO 10-17 21:10:28 [core.py:730] Waiting for init message from front-end.
(EngineCore_DP0 pid=782659) INFO 10-17 21:10:28 [core.py:97] Initializing a V1 LLM engine (v0.11.1rc2.dev115+g3aeb19a39) with config: model='Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound', speculative_config=None, tokenizer='Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.float16, max_seq_len=3072, download_dir='/media/ninemeow/Kioxia', load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=True, quantization=auto-round, enforce_eager=True, kv_cache_dtype=auto, device_config=xpu, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser=''), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None), seed=0, served_model_name=Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound, enable_prefix_caching=False, chunked_prefill_enabled=True, pooler_config=None, compilation_config={'level': None, 'mode': 0, 'debug_dump_path': None, 'cache_dir': '', 'backend': 'inductor', 'custom_ops': ['all'], 'splitting_ops': None, 'use_inductor': None, 'compile_sizes': [], 'inductor_compile_config': {'enable_auto_functionalized_v2': False}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.NONE: 0>, 'use_cudagraph': True, 'cudagraph_num_of_warmups': 0, 'cudagraph_capture_sizes': None, 'cudagraph_copy_inputs': False, 'full_cuda_graph': False, 'use_inductor_graph_partition': False, 'pass_config': {}, 'max_capture_size': None, 'local_cache_dir': None}
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
[Gloo] Rank 0 is connected to 0 peer ranks. Expected number of connected peer ranks is : 0
(EngineCore_DP0 pid=782659) INFO 10-17 21:10:29 [parallel_state.py:1325] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, TP rank 0, EP rank 0
2025:10:17-21:10:29:(782659) |CCL_WARN| value of CCL_ATL_TRANSPORT changed to be ofi (default:mpi)
2025:10:17-21:10:29:(782659) |CCL_WARN| could not get local_idx/count from environment variables, trying to get them from ATL
2025:10:17-21:10:29:(782659) |CCL_WARN| device_family is unknown, topology discovery could be incorrect, it might result in suboptimal performance
2025:10:17-21:10:29:(782659) |CCL_WARN| Applying sycl-kernels, but device family is not recognized
(EngineCore_DP0 pid=782659) INFO 10-17 21:10:29 [gpu_model_runner.py:2843] Starting to load model Intel/Qwen3-Next-80B-A3B-Instruct-int4-AutoRound...
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] EngineCore failed to start.
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] Traceback (most recent call last):
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core.py", line 784, in run_engine_core
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] engine_core = EngineCoreProc(*args, **kwargs)
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core.py", line 552, in init
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] super().init(
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core.py", line 106, in init
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] self.model_executor = executor_class(vllm_config)
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/executor/executor_base.py", line 54, in init
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] self._init_executor()
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/executor/uniproc_executor.py", line 48, in _init_executor
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] self.collective_rpc("load_model")
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/executor/uniproc_executor.py", line 74, in collective_rpc
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] return [run_method(self.driver_worker, method, args, kwargs)]
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/utils/init.py", line 2089, in run_method
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] return func(*args, **kwargs)
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/worker/gpu_worker.py", line 229, in load_model
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] self.model_runner.load_model(eep_scale_up=eep_scale_up)
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/worker/gpu_model_runner.py", line 2873, in load_model
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] self.model = model_loader.load_model(
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/model_loader/base_loader.py", line 49, in load_model
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] model = initialize_model(
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/model_loader/utils.py", line 65, in initialize_model
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] return model_class(vllm_config=vllm_config, prefix=prefix)
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/models/qwen3_next.py", line 1118, in init
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] self.model = Qwen3NextModel(
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/compilation/decorators.py", line 234, in init
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] old_init(self, vllm_config=vllm_config, prefix=prefix, **kwargs)
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/models/qwen3_next.py", line 953, in init
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] self.start_layer, self.end_layer, self.layers = make_layers(
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/models/utils.py", line 640, in make_layers
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] maybe_offload_to_cpu(layer_fn(prefix=f"{prefix}.{idx}"))
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/models/utils.py", line 599, in maybe_offload_to_cpu
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] p.data = get_cuda_view_from_cpu_tensor(cpu_data)
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/utils/init.py", line 1553, in get_cuda_view_from_cpu_tensor
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] return torch.ops._C.get_cuda_view_from_cpu_tensor(cpu_tensor)
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/torch/_ops.py", line 1353, in getattr
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] raise AttributeError(
(EngineCore_DP0 pid=782659) ERROR 10-17 21:10:29 [core.py:793] AttributeError: '_OpNamespace' '_C' object has no attribute 'get_cuda_view_from_cpu_tensor'
(EngineCore_DP0 pid=782659) Process EngineCore_DP0:
(EngineCore_DP0 pid=782659) Traceback (most recent call last):
(EngineCore_DP0 pid=782659) File "/home/ninemeow/.pyenv/versions/3.12.9/lib/python3.12/multiprocessing/process.py", line 314, in _bootstrap
(EngineCore_DP0 pid=782659) self.run()
(EngineCore_DP0 pid=782659) File "/home/ninemeow/.pyenv/versions/3.12.9/lib/python3.12/multiprocessing/process.py", line 108, in run
(EngineCore_DP0 pid=782659) self._target(*self._args, **self._kwargs)
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core.py", line 797, in run_engine_core
(EngineCore_DP0 pid=782659) raise e
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core.py", line 784, in run_engine_core
(EngineCore_DP0 pid=782659) engine_core = EngineCoreProc(*args, **kwargs)
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core.py", line 552, in init
(EngineCore_DP0 pid=782659) super().init(
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core.py", line 106, in init
(EngineCore_DP0 pid=782659) self.model_executor = executor_class(vllm_config)
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/executor/executor_base.py", line 54, in init
(EngineCore_DP0 pid=782659) self._init_executor()
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/executor/uniproc_executor.py", line 48, in _init_executor
(EngineCore_DP0 pid=782659) self.collective_rpc("load_model")
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/executor/uniproc_executor.py", line 74, in collective_rpc
(EngineCore_DP0 pid=782659) return [run_method(self.driver_worker, method, args, kwargs)]
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/utils/init.py", line 2089, in run_method
(EngineCore_DP0 pid=782659) return func(*args, **kwargs)
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/worker/gpu_worker.py", line 229, in load_model
(EngineCore_DP0 pid=782659) self.model_runner.load_model(eep_scale_up=eep_scale_up)
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/worker/gpu_model_runner.py", line 2873, in load_model
(EngineCore_DP0 pid=782659) self.model = model_loader.load_model(
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/model_loader/base_loader.py", line 49, in load_model
(EngineCore_DP0 pid=782659) model = initialize_model(
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/model_loader/utils.py", line 65, in initialize_model
(EngineCore_DP0 pid=782659) return model_class(vllm_config=vllm_config, prefix=prefix)
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/models/qwen3_next.py", line 1118, in init
(EngineCore_DP0 pid=782659) self.model = Qwen3NextModel(
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/compilation/decorators.py", line 234, in init
(EngineCore_DP0 pid=782659) old_init(self, vllm_config=vllm_config, prefix=prefix, **kwargs)
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/models/qwen3_next.py", line 953, in init
(EngineCore_DP0 pid=782659) self.start_layer, self.end_layer, self.layers = make_layers(
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/models/utils.py", line 640, in make_layers
(EngineCore_DP0 pid=782659) maybe_offload_to_cpu(layer_fn(prefix=f"{prefix}.{idx}"))
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/model_executor/models/utils.py", line 599, in maybe_offload_to_cpu
(EngineCore_DP0 pid=782659) p.data = get_cuda_view_from_cpu_tensor(cpu_data)
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/utils/init.py", line 1553, in get_cuda_view_from_cpu_tensor
(EngineCore_DP0 pid=782659) return torch.ops._C.get_cuda_view_from_cpu_tensor(cpu_tensor)
(EngineCore_DP0 pid=782659) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=782659) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/torch/_ops.py", line 1353, in getattr
(EngineCore_DP0 pid=782659) raise AttributeError(
(EngineCore_DP0 pid=782659) AttributeError: '_OpNamespace' '_C' object has no attribute 'get_cuda_view_from_cpu_tensor'
(APIServer pid=782393) Traceback (most recent call last):
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/bin/vllm", line 33, in
(APIServer pid=782393) sys.exit(load_entry_point('vllm==0.11.1rc2.dev115+g3aeb19a39.xpu', 'console_scripts', 'vllm')())
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/entrypoints/cli/main.py", line 73, in main
(APIServer pid=782393) args.dispatch_function(args)
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/entrypoints/cli/serve.py", line 62, in cmd
(APIServer pid=782393) uvloop.run(run_server(args))
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/uvloop/init.py", line 96, in run
(APIServer pid=782393) return __asyncio.run(
(APIServer pid=782393) ^^^^^^^^^^^^^^
(APIServer pid=782393) File "/home/ninemeow/.pyenv/versions/3.12.9/lib/python3.12/asyncio/runners.py", line 195, in run
(APIServer pid=782393) return runner.run(main)
(APIServer pid=782393) ^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/home/ninemeow/.pyenv/versions/3.12.9/lib/python3.12/asyncio/runners.py", line 118, in run
(APIServer pid=782393) return self._loop.run_until_complete(task)
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "uvloop/loop.pyx", line 1518, in uvloop.loop.Loop.run_until_complete
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/uvloop/init.py", line 48, in wrapper
(APIServer pid=782393) return await main
(APIServer pid=782393) ^^^^^^^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/entrypoints/openai/api_server.py", line 1920, in run_server
(APIServer pid=782393) await run_server_worker(listen_address, sock, args, **uvicorn_kwargs)
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/entrypoints/openai/api_server.py", line 1936, in run_server_worker
(APIServer pid=782393) async with build_async_engine_client(
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/home/ninemeow/.pyenv/versions/3.12.9/lib/python3.12/contextlib.py", line 210, in aenter
(APIServer pid=782393) return await anext(self.gen)
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/entrypoints/openai/api_server.py", line 191, in build_async_engine_client
(APIServer pid=782393) async with build_async_engine_client_from_engine_args(
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/home/ninemeow/.pyenv/versions/3.12.9/lib/python3.12/contextlib.py", line 210, in aenter
(APIServer pid=782393) return await anext(self.gen)
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/entrypoints/openai/api_server.py", line 238, in build_async_engine_client_from_engine_args
(APIServer pid=782393) async_llm = AsyncLLM.from_vllm_config(
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/utils/functools.py", line 116, in inner
(APIServer pid=782393) return fn(*args, **kwargs)
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/async_llm.py", line 210, in from_vllm_config
(APIServer pid=782393) return cls(
(APIServer pid=782393) ^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/async_llm.py", line 132, in init
(APIServer pid=782393) self.engine_core = EngineCoreClient.make_async_mp_client(
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core_client.py", line 121, in make_async_mp_client
(APIServer pid=782393) return AsyncMPClient(*client_args)
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core_client.py", line 807, in init
(APIServer pid=782393) super().init(
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/core_client.py", line 468, in init
(APIServer pid=782393) with launch_core_engines(vllm_config, executor_class, log_stats) as (
(APIServer pid=782393) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=782393) File "/home/ninemeow/.pyenv/versions/3.12.9/lib/python3.12/contextlib.py", line 144, in exit
(APIServer pid=782393) next(self.gen)
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/utils.py", line 816, in launch_core_engines
(APIServer pid=782393) wait_for_engine_startup(
(APIServer pid=782393) File "/media/ninemeow/Kioxia/vllm/venv/lib/python3.12/site-packages/vllm-0.11.1rc2.dev115+g3aeb19a39.xpu-py3.12.egg/vllm/v1/engine/utils.py", line 873, in wait_for_engine_startup
(APIServer pid=782393) raise RuntimeError(
(APIServer pid=782393) RuntimeError: Engine core initialization failed. See root cause above. Failed core proc(s): {}
Is it not implemented?
Versions
=====================================
PyTorch version: 2.8.0+xpu
PyTorch CXX11 ABI: Yes
IPEX version: 2.8.10.post0+xpu
IPEX commit: 48933bfea
Build type: Release
OS: LMDE 7 (gigi) (x86_64)
GCC version: (Debian 14.2.0-19) 14.2.0
Clang version: N/A
IGC version: 2025.1.1 (2025.1.1.20250418)
CMake version: version 4.1.0
Libc version: glibc-2.41
Python version: 3.12.9 (main, Apr 5 2025, 12:59:17) [GCC 12.2.0] (64-bit runtime)
Python platform: Linux-6.16.3+deb13-amd64-x86_64-with-glibc2.41
Is XPU available: True
DPCPP runtime: 2025.1
MKL version: 2025.1
GPU models and configuration onboard:
- Intel(R) Arc(TM) A770 Graphics
GPU models and configuration detected:
- [0] _XpuDeviceProperties(name='Intel(R) Arc(TM) A770 Graphics', platform_name='Intel(R) oneAPI Unified Runtime over Level-Zero', type='gpu', driver_version='1.6.35096+9', total_memory=15473MB, max_compute_units=512, gpu_eu_count=512, gpu_subslice_count=32, max_work_group_size=1024, max_num_sub_groups=128, sub_group_sizes=[8 16 32], has_fp16=1, has_fp64=0, has_atomic64=1)
Driver version:
- intel_opencl: 25.35.35096.9-0
- level_zero: N/A
CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 39 bits physical, 48 bits virtual
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: GenuineIntel
Model name: 13th Gen Intel(R) Core(TM) i5-13490F
CPU family: 6
Model: 191
Thread(s) per core: 2
Core(s) per socket: 10
Socket(s): 1
Stepping: 2
CPU(s) scaling MHz: 38%
CPU max MHz: 4800.0000
CPU min MHz: 800.0000
BogoMIPS: 4992.00
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rdseed adx smap clflushopt clwb intel_pt sha_ni xsaveopt xsavec xgetbv1 xsaves split_lock_detect user_shstk avx_vnni dtherm ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp hwp_pkg_req hfi vnmi umip pku ospke waitpkg gfni vaes vpclmulqdq rdpid movdiri movdir64b fsrm md_clear serialize arch_lbr ibt flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 416 KiB (10 instances)
L1i cache: 448 KiB (10 instances)
L2 cache: 9.5 MiB (7 instances)
L3 cache: 24 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Ghostwrite: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Old microcode: Not affected
Vulnerability Reg file data sampling: Mitigation; Clear Register File
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Enhanced / Automatic IBRS; IBPB conditional; PBRSB-eIBRS SW sequence; BHI BHI_DIS_S
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Versions of relevant libraries:
[pip] dpcpp-cpp-rt==2025.1.1
[pip] impi-rt==2021.15.0
[pip] intel-cmplr-lib-rt==2025.1.1
[pip] intel-cmplr-lib-ur==2025.1.1
[pip] intel-cmplr-lic-rt==2025.1.1
[pip] intel_extension_for_pytorch==2.8.10.post0+xpu
[pip] intel-opencl-rt==2025.1.1
[pip] intel-openmp==2025.1.1
[pip] intel-pti==0.12.3
[pip] intel-sycl-rt==2025.1.1
[pip] mkl==2025.1.0
[pip] numpy==2.2.6
[pip] oneccl==2021.15.2
[pip] oneccl-devel==2021.15.2
[pip] onemkl-sycl-blas==2025.1.0
[pip] onemkl-sycl-dft==2025.1.0
[pip] onemkl-sycl-lapack==2025.1.0
[pip] onemkl-sycl-rng==2025.1.0
[pip] onemkl-sycl-sparse==2025.1.0
[pip] pytorch-triton-xpu==3.4.0
[pip] torch==2.8.0+xpu
[pip] torchaudio==2.8.0+xpu
[pip] torchvision==0.23.0+xpu
[pip] transformers==4.57.1