search

vbelz / Speech-enhancement

Deep learning for audio denoising
MIT License
620 stars 124 forks source link

Question on Error of Invalid Instruction (core dumped) #11

Open ashwinijha6 opened 3 years ago

ashwinijha6 commented 3 years ago

Hey @vbelz I had a question : While running the python main.py --mode='data_creation' I get the error Invalid Instruction (core dumped) I guess it because of the tensorflow version (1.15.2) as my cpu does not support AVX But it would not give an error if I use tensorflow version (1.5)

If I want to use the same version of tensorflow that is 1.15.2 what could be an alternative?

vbelz commented 3 years ago

Hey Ashwini, Not sure why you get this error. You could switch to version 1.15 if it works.

Thanks

Vincent

Le mar. 26 janv. 2021 à 07:59, Ashwini Jha notifications@github.com a écrit :

Hey @vbelz https://github.com/vbelz I had a question : While running the python main.py --mode='data_creation' I get the error Invalid Instruction (core dumped) I guess it because of the tensorflow version (1.15.2) as my cpu does not support AVX But it would not give an error if I use tensorflow version (1.5)

If I want to use the same version of tensorflow that is 1.15.2 what could be an alternative?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/vbelz/Speech-enhancement/issues/11, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJHCRMFIV3V4UXRN24DIMNLS32OBBANCNFSM4WTIFG7Q .

ashwinijha6 commented 3 years ago

Hey Ashwini, Not sure why you get this error. You could switch to version 1.15 if it works. Thanks Vincent Le mar. 26 janv. 2021 à 07:59, Ashwini Jha notifications@github.com a écrit : … Hey @vbelz https://github.com/vbelz I had a question : While running the python main.py --mode='data_creation' I get the error Invalid Instruction (core dumped) I guess it because of the tensorflow version (1.15.2) as my cpu does not support AVX But it would not give an error if I use tensorflow version (1.5) If I want to use the same version of tensorflow that is 1.15.2 what could be an alternative? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#11>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJHCRMFIV3V4UXRN24DIMNLS32OBBANCNFSM4WTIFG7Q .

Sorry @vbelz it is Illegal instruction (core dumped) & I guess it is because my cpu does not support avx. Tensorflow version higher than 1.5 need support of avx. Check this out once https://tech.amikelive.com/node-887/how-to-resolve-error-illegal-instruction-core-dumped-when-running-import-tensorflow-in-a-python-program/

ashwinijha6 commented 3 years ago

hey @vbelz I just resolved the error by downgrading the tensorflow version to 1.5 & librosa to 0.7.2 & numba to 0.48 Just wanted to know I have used .wav noise files (400 in number) & .flac voice files (402 in number) & after running python main.py --mode=='data_creation' I get the spectrograms in .npy format (as shown in the screenshots) would it be good for training. What do you recommend? Should I increase the number of frames to 40000 as you mentioned in the README.md spectrogram

jacktheripper-17 commented 3 years ago

@ashwinijha6 Hi, are you getting any issues in the prediction stage? Especially when giving your own file as an argument in the command line?

ashwinijha6 commented 3 years ago

@ashwinijha6 Hi, are you getting any issues in the prediction stage? Especially when giving your own file as an argument in the command line?

Hey @jacktheripper-17 I just used the command line for data creation using data creation mode but for training & prediction I used colab notebook which is available in the project here taking the .npy files as data that was created.