PytorchBackend.to_sharedarr#
- PytorchBackend.to_sharedarr(arr, shared_memory_handler=None)[source]#
Converts an array to an object shared in memory. The output of this function will only be passed to
from_sharedarr()
, hence the return values can be modified in particular backends to match the expected input data.- Parameters:
- arrBackendArray
Numpy array to convert.
- shared_memory_handlertype, optional
The type of shared memory handler. Default is None.
- Returns:
- Tuple[shared_memory.SharedMemory, tuple of ints, dtype]
The shared memory object containing the numpy array, its shape and dtype.