Is towhee and the models support the 'npu' device,the device is from huawei atlas 310P

[sportzhang]

i want use the clip4clip video process pipeline in 'npu' rather than 'cuda' device,but got the error:
EE8888: Inner Error, Please contact support engineer!
ctx is NULL![FUNC:GetDevErrMsg][FILE:api_impl.cc][LINE:3490]
rtGetDevMsg execute failed, reason=[context pointer null][FUNC:FuncErrorReason][FILE:error_message_manage.cc][LINE:49]
, Traceback (most recent call last):
File "/root/anaconda3/lib/python3.7/site-packages/towhee/runtime/nodes/node.py", line 158, in _call
return True, self._op(*inputs), None
File "/root/.towhee/operators/video-text-embedding/clip4clip/versions/main/clip4clip.py", line 65, in call
vec = self._inference_from_video(data)
File "/root/.towhee/operators/video-text-embedding/clip4clip/versions/main/clip4clip.py", line 94, in _inference_from_video
video = torch.as_tensor(video).float().to(self.device)
File "/root/anaconda3/lib/python3.7/site-packages/torch_npu/utils/device_guard.py", line 37, in wrapper
return func(*args, **kwargs)
File "/root/anaconda3/lib/python3.7/site-packages/torch_npu/utils/tensor_methods.py", line 82, in to
return torch_npu.C.to(self, *args, **kwargs)
RuntimeError: allocate:/usr1/workspace/FPTA_Daily_Plugin_open_date/CODE/torch_npu/csrc/core/npu/NPUCachingAllocator.cpp:1061 NPU error, error code is 107002
EE8888: Inner Error, Please contact support engineer!
ctx is NULL![FUNC:GetDevErrMsg][FILE:api_impl.cc][LINE:3490]
rtGetDevMsg execute failed, reason=[context pointer null][FUNC:FuncErrorReason][FILE:error_message_manage.cc][LINE:49]

[junjiejiangjjj]

hi @sportzhang , this error seems to be a torch_npu error. Towhee is not aware of the specific device and simply passes the information to pytorch.