load audio files server and local path issues

Start by tracing AudioTable.load_audio_files with the shown local_audio_path and server_audio_path, then compare its behavior with the table.loadtable call and the mycaslib configuration. Reproduce the PATH/DOCUMENT error using the provided paths; done means the intended audio subset loads into its own table without that error.

modelPath = '/srv/nfs/kubedata/cas-landingzone/sbxtot/sounds'

s.addcaslib(activeonadd=True,datasource={'srctype':'path'},name='mycaslib',path=modelPath,subdirectories=True)
s.setsessopt(caslib='mycaslib')

local_data_test = '/srv/nfs/kubedata/cas-landingzone/sbxtot/sounds/test'
server_data_test = '/srv/nfs/kubedata/cas-landingzone/sbxtot/sounds/test'
local_data_train = '/srv/nfs/kubedata/cas-landingzone/sbxtot/sounds/train'
server_data_train = '/srv/nfs/kubedata/cas-landingzone/sbxtot/sounds/train'
local_data_validate = '/srv/nfs/kubedata/compute-landingzone/sbxtot/sounds/validate'
server_data_validate = 'srv/nfs/kubedata/compute-landingzone/sbxtot/sounds/validate'

audio_table_train = AudioTable.load_audio_files(s,
casout={'name':'audio_files_train','replace':'True'},
local_audio_path=local_data_train,
server_audio_path=server_playground_data)

I can't understand why this procedure can't process those files and constantly throw that error:

Cannot convert file /srv/nfs/kubedata/cas-landingzone/sbxtot/sounds/train/abnormal/00000110.wav
Number of files processed: 0
File conversions are finished.
ERROR: A PATH type caslib is required when using filetype DOCUMENT.
ERROR: The action stopped due to errors.
ERROR: A PATH type caslib is required when using filetype DOCUMENT.
ERROR: The action stopped due to errors.

While in the meantime, the following procedure works and create my table but it includes all the subdirectories of sounds (test, train, validate) content in one file

rt2 = s.retrieve('table.loadtable', _messagelevel='error', casout= {'name':'audio_files_train','replace':'True'},
caslib='mycaslib', path='',
importOptions=dict(fileType='AUDIO', contents=True, recurse=True))

Can you let me know how to solve this issue ?
Thanks,
Tom.