Hugging Face's logo

crlandsc
/
bsrnn-drums

audio source separation
music demixing
band-split recurrent neural network
bsrnn
spectrogram
drums
Model card Files Community
1

Is this model corrupted?

#1
by filipfranek - opened
Discussion
filipfranek
 
root@d5f815dce7a8:/src# python3 inference.py -i inputs/audio_example.wav -o outputs -t other --ckpt-path saved_models/other/other.ckpt
Traceback (most recent call last):
  File "/src/inference.py", line 111, in <module>
    main(args)
  File "/src/inference.py", line 64, in main
    program = InferenceProgram(**args)
  File "/src/inference.py", line 45, in __init__
    self.sep = Separator(self.cfg, self.ckpt_path)
  File "/src/separator.py", line 25, in __init__
    self.model = self.initialize_modules()
  File "/src/separator.py", line 55, in initialize_modules
    state_dict = load_pl_state_dict(self.ckpt_path, device='cpu')
  File "/src/utils/utils_inference.py", line 14, in load_pl_state_dict
    sd = torch.load(path, map_location=device)
  File "/usr/local/lib/python3.10/site-packages/torch/serialization.py", line 789, in load
    return _load(opened_zipfile, map_location, pickle_module, **pickle_load_args)
  File "/usr/local/lib/python3.10/site-packages/torch/serialization.py", line 1131, in _load
    result = unpickler.load()
  File "/usr/local/lib/python3.10/pickle.py", line 1213, in load
    dispatch[key[0]](self)
  File "/usr/local/lib/python3.10/pickle.py", line 1254, in load_binpersid
    self.append(self.persistent_load(pid))
  File "/usr/local/lib/python3.10/site-packages/torch/serialization.py", line 1101, in persistent_load
    load_tensor(dtype, nbytes, key, _maybe_decode_ascii(location))
  File "/usr/local/lib/python3.10/site-packages/torch/serialization.py", line 1079, in load_tensor
    storage = zip_file.get_storage_from_record(name, numel, torch.UntypedStorage).storage().untyped()
RuntimeError: PytorchStreamReader failed reading file data/78: invalid header or archive is corrupted
minizero

I'm getting this with the drums ckpt as well.
python inference.py -i sample.mp3 -o outputs -t drums --ckpt-path saved_models/drums/drums.ckpt ─╯
Traceback (most recent call last):
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/src/inference.py", line 111, in
main(args)
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/src/inference.py", line 64, in main
program = InferenceProgram(**args)
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/src/inference.py", line 45, in init
self.sep = Separator(self.cfg, self.ckpt_path)
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/src/separator.py", line 25, in init
self.model = self.initialize_modules()
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/src/separator.py", line 55, in initialize_modules
state_dict = load_pl_state_dict(self.ckpt_path, device='cpu')
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/src/utils/utils_inference.py", line 14, in load_pl_state_dict
sd = torch.load(path, map_location=device)
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/.env/lib/python3.9/site-packages/torch/serialization.py", line 789, in load
return _load(opened_zipfile, map_location, pickle_module, **pickle_load_args)
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/.env/lib/python3.9/site-packages/torch/serialization.py", line 1131, in _load
result = unpickler.load()
File "/usr/local/Cellar/[email protected]/3.9.18_1/Frameworks/Python.framework/Versions/3.9/lib/python3.9/pickle.py", line 1212, in load
dispatchkey[0]
File "/usr/local/Cellar/[email protected]/3.9.18_1/Frameworks/Python.framework/Versions/3.9/lib/python3.9/pickle.py", line 1253, in load_binpersid
self.append(self.persistent_load(pid))
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/.env/lib/python3.9/site-packages/torch/serialization.py", line 1101, in persistent_load
load_tensor(dtype, nbytes, key, _maybe_decode_ascii(location))
File "/Users/amir/Projects/Music-Demixing-with-Band-Split-RNN/.env/lib/python3.9/site-packages/torch/serialization.py", line 1079, in load_tensor
storage = zip_file.get_storage_from_record(name, numel, torch.UntypedStorage).storage().untyped()
RuntimeError: PytorchStreamReader failed reading file data/76: invalid header or archive is corrupted

crlandsc
 Owner

@filipfranek @minizero Sorry for the delayed response. I am not sure what happened with those checkpoints, but they also gave me the same errors when I downloaded and tried them. I re-uploaded fixed checkpoints for both drums and other. The existing vocals and bass checkpoints did not appear corrupted. So I believe that all models should work now, but let me know if you run into more issues. Cheers!

Sign up or log in to comment