I couldn't find a nice high-level package that is fast enough to serialize and deserialize numpy arrays while preserving the type and shape informations.

pip install numpy-serializer 

import numpy_serializer as ns import numpy as np a = np.random.normal(size=(50,120,150)) b = ns.to_bytes(a) c = ns.from_bytes(b) assert np.array_equal(a,c)

This project is licensed under the GPL-3 license.