clovaai / voxceleb_trainer

In defence of metric learning for speaker recognition
MIT License
1.03k stars 272 forks source link

ValueError: source code string cannot contain null bytes #56

Closed dendi123 closed 4 years ago

dendi123 commented 4 years ago

Hi, Since I'm kinda new to VGGVox and Python, I'm currently trying to train VGGVox Model but I have some errors when running the trainning code python ./trainSpeakerNet.py --model ResNetSE34L --log_input True --encoder SAP --trainfunc amsoftmax --save_path ./data/exp1 --nSpeakers 5994 --batch_size 200 --scale 30 --margin 0.3 --train_list ./train_list.txt --test_list ./test_list.txt --train_path ./voxceleb/voxceleb2 --test_path ./voxceleb/voxceleb1

Actual Behavior Traceback (most recent call last): File "./trainSpeakerNet.py", line 93, in s = SpeakerNet(**vars(args)); File "D:\VGGvox\VGGvoxtrainer\voxceleb_trainer\SpeakerNet.py", line 19, in init SpeakerNetModel = importlib.import_module('models.'+model).getattribute('MainModel') File "G:\Anaconda\lib\importlib__init__.py", line 127, in import_module return _bootstrap._gcd_import(name[level:], package, level) File "", line 1006, in _gcd_import File "", line 983, in _find_and_load File "", line 967, in _find_and_load_unlocked File "", line 677, in _load_unlocked File "", line 728, in exec_module File "", line 219, in _call_with_frames_removed File "D:\VGGvox\VGGvoxtrainer\voxceleb_trainer\models\VGGVox.py", line 5, in import torchaudio ValueError: source code string cannot contain null bytes

I got this errors with ResNetSE34L and VGGVox a different error shown up with ResNetSE34 but I think I can handle that Your System configuration Python version: 3.5.6 torch 1.6.0 torchaudio 0.6.0

Thanks, Wish you have a good day

joonson commented 4 years ago

Hi, Windows platform is not supported.

huyvl commented 4 years ago

Hi, Windows platform is not supported.

Oh thanks you. Very appreciate that