How to execute LisGAN.

[naumanbashir70]

Hi, @lijin118

I am trying to run your program but it is giving me error continuously. I have run LisGAN.py but is asked for dataset, I have downloaded dataset and properly set it path in your code. Now, it is giving h5py error. Does it need conversion from .MAT file to .hdf5?

Thanks

[HuangYL-0619]

https://github.com/naumanbashir70 Hello! when I run this program,I find it can't run succeedly,Do you have such problem like
File "F:/PythonProject/try0/lisgan.py", line 262, in
sample()
File "F:/PythonProject/try0/lisgan.py", line 135, in sample
input_att.copy_(batch_att)
RuntimeError: The size of tensor a (1024) must match the size of tensor b (85) at non-singleton dimension 1

[naumanbashir70]

@yuran1024
Which dataset you are using?

[HuangYL-0619]

@naumanbashir70 I downloaded datasets through its link.When I run this program using AWA1、APY、SUN、or CUB,it has the same problem.The problem happened in function “sample” in lisgan.py.Do you have some advice?

[naumanbashir70]

Are you running the LigGAN.py directory or with the parameters.
I was receiving this error.
KeyError : Could not found train_loc . as the dataset was changed.

[HuangYL-0619]

@naumanbashir70 I have solved this problem.The name in the file doesn't match the name in the code.You can read "README" and find its name are "att" and "original_att".

[naumanbashir70]

Your code is running now?

[HuangYL-0619]

@naumanbashir70 No,I have a new problem.The tenosr doesn't match.I don't know how to debug it.Do you have some advice?

[naumanbashir70]

Send me the complete screenshot of error.

[HuangYL-0619]

@naumanbashir70

[naumanbashir70]

Changes the att size in lisgan.py according to your dataset.

[naumanbashir70]

Send me your
lisga.py and util.py
Thanks

[HuangYL-0619]

@naumanbashir70
try0.zip

Thank you for your help!

[naumanbashir70]

Try this
lisgan.zip

[HuangYL-0619]

@naumanbashir70 Thank for you help!You can run your code now? I have a new problem.I use python3.5 + pytorch1.4 +cuda8.0.I debug it by what it tips.

But it happens another pronlem

[zhaizhibo]

@naumanbashir70 Thank for you help!You can run your code now? I have a new problem.I use python3.5 + pytorch1.4 +cuda8.0.I debug it by what it tips.

But it happens another pronlem

I have the same problem now. Have you solved that problem?