DaphneLib supports efficiently exchanging data with numpy via shared memory. However, at the moment, only numpy arrays with 1 or 2 dimensions are supported. Arrays with more than 2 dimensions should be supported as well. Inspiration can be taken from the data exchange with TensorFlow and PyTorch, where tensors with more than 2 dimensions are already supported.
DaphneLib supports efficiently exchanging data with numpy via shared memory. However, at the moment, only numpy arrays with 1 or 2 dimensions are supported. Arrays with more than 2 dimensions should be supported as well. Inspiration can be taken from the data exchange with TensorFlow and PyTorch, where tensors with more than 2 dimensions are already supported.