How can we paas model directory for Spleeter to split audios?

I am using Spleeter to split audios into vocals and accompaniment. Upon executing the below code, I noticed that it downloads a model file everytime i execute the code. Is there any way to paas the model directory as a parameter in the code?

**Code**

def SplitFile(AudioFile,OutputPath):
model='spleeter:2stems'
if not os.path.exists(OutputPath):
os.mkdir(OutputPath)
try:
separator = Separator(model)
separator.separate_to_file(AudioFile,OutputPath,duration=10800)
print('Vocals and the music are separated and stored in the directory',flush=True)
return 'successfully splitted'
except Exception as e:
print('error in spleeter',e,flush=True)
return None

Start with the shown `Separator(model)` call and read the `Separator` usage or API documentation to see whether a model directory can be supplied. The issue names no files or tests; done means clarifying the supported way to reuse a model directory, or documenting that this is not supported.

Written by the indexing model from the issue text.