Export model
-
Adds export module for neural networks. In its current form, this new module contains only one function:
export_to_pamguard
. - Also added two tests of this new function.
- Extended the interface of the nn_interface.save_model method to allow user to specify audio representation as either dict or path to json.
-
Implemented new
encode_audio_representation
method indata_handling.parsing
which essentially is the inverse mapping of the existingparse_audio_representation
function -
Modified the behaviour of the
parse_audio_representation
function as suggested in #109 (closed)