香橙派运行python3 -c "import torch;import torch_npu; a = torch.randn(3, 4).npu(); print(a + a);"时发生报错。

将7.0.0对应的kernel包安装之后依然是相同的报错。

环境：

driver == 23.0.0

cann == 7.0.0

torch == 2.1.0

torch_npu == 2.1.0.post2+git64bdab5

 

(base) root@orangepiaipro:/home# python3 -c "import torch;import torch_npu; a = torch.randn(3, 4).npu(); print(a + a);" 

/usr/local/miniconda3/lib/python3.9/site-packages/torch_npu/dynamo/__init__.py:18: UserWarning: Register eager implementation for the 'npu' backend of dynamo, as torch_npu was not compiled with torchair. 

  warnings.warn( 

Killed 

(base) root@orangepiaipro:/home# Process ForkServerPoolWorker-4: 

Process ForkServerPoolWorker-8: 

Process ForkServerPoolWorker-7: 

Process ForkServerPoolWorker-2: 

Process ForkServerPoolWorker-3: 

Process ForkServerPoolWorker-6: 

Traceback (most recent call last): 

  File "/usr/local/miniconda3/lib/python3.9/multiprocessing/pool.py", line 131, in worker 

    put((job, i, result)) 

  File "/usr/local/miniconda3/lib/python3.9/multiprocessing/queues.py", line 378, in put 

    self._writer.send_bytes(obj) 

  File "/usr/local/miniconda3/lib/python3.9/multiprocessing/connection.py", line 205, in send_bytes 

    self._send_bytes(m[offset:offset + size]) 

  File "/usr/local/miniconda3/lib/python3.9/multiprocessing/connection.py", line 416, in _send_bytes 

    self._send(header + buf) 

  File "/usr/local/miniconda3/lib/python3.9/multiprocessing/connection.py", line 373, in _send 

    n = write(self._handle, buf) 

BrokenPipeError: [Errno 32] Broken pipe 

 

…………

/usr/local/miniconda3/lib/python3.9/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 97 leaked semaphore objects to clean up at shutdown 

  warnings.warn('resource_tracker: There appear to be %d '