The types produced by the library can be used by other libraries that accept numpy arrays (and in future, tensors).
However, it would be good to add in dedicated support/fast path to the likes of:
- plotting
- serialization/deserialization
- more general I/O (e.g. spectrogram.save(path, ...?))
The types produced by the library can be used by other libraries that accept numpy arrays (and in future, tensors).
However, it would be good to add in dedicated support/fast path to the likes of: