qwen_vllm | sub_out = torch.nn.functional.scaled_dot_product_attention( qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] max(): Expected reduction dim 2 to have non-zero size. qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] Traceback (most recent call last): qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 464, in run_mp_engine qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] engine = MQLLMEngine.from_vllm_config( qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 139, in from_vllm_config qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] return cls( qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 93, in __init__ qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] self.engine = LLMEngine(*args, **kwargs) qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/llm_engine.py", line 276, in __init__ qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] self._initialize_kv_caches() qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/llm_engine.py", line 424, in _initialize_kv_caches qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] self.model_executor.determine_num_available_blocks()) qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/executor/executor_base.py", line 104, in determine_num_available_blocks qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] results = self.collective_rpc("determine_num_available_blocks") qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/executor/executor_base.py", line 332, in collective_rpc qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] return self._run_workers(method, *args, **(kwargs or {})) qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/executor/mp_distributed_executor.py", line 186, in _run_workers qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] driver_worker_output = run_method(self.driver_worker, sent_method, qwen_vllm | ERROR 12-30 17:29:33 [engine.py:476] File "/usr/local/lib/python3.10/dist-packages/vllm/utils/__init__.py", line 2910, in run_method 这是大概的错误详情