ValueError: invalid literal for int()

[jolitte]

I have tried the python code based on the README.MD file
It encodes the input.png file successfully but the decode gets an error.
ValueError: invalid literal for int()
Need help to solve this issue and how to use the steganogan properly.
Thank you in advance.

Below is the python code:

from steganogan import SteganoGAN
steganogan = SteganoGAN.load(architecture='dense')

steganogan.encode('research/Input.png', 'research/Output.png', 'This is a super secret message!')
steganogan.decode('research/Output.png')
'This is a super secret message!'

[pvk-developer]

Thanks for reporting this @jolitte !

Would you mind reporting more information about your environment such as python version, operating system and the output of a pip freeze command ?

I'll see if I can reproduce this on my side too.

[DG-Abraham]

I met the same problem such as @jolitte, when i try to decode it report the error which signs " ValueError: invalid literal for int() with base 2: 'TrueFalseTrueTrueTrueTrueTrueFalse' ". I work this project in Windows 10 which installed python 3.8 and pytorch 1.8.1 based on cuda 11.
Then I turn to use pytorch-cpu 1.0.0, it works but the speed of encoding and decoding is too low.

[pvk-developer]

I met the same problem such as @jolitte, when i try to decode it report the error which signs " ValueError: invalid literal for int() with base 2: 'TrueFalseTrueTrueTrueTrueTrueFalse' ". I work this project in Windows 10 which installed python 3.8 and pytorch 1.8.1 based on cuda 11.
Then I turn to use pytorch-cpu 1.0.0, it works but the speed of encoding and decoding is too low.

@DG-Abraham We don't support python 3.8 officially and the torch dependency is set to 1.0.0 as we solved the issue #42

[DG-Abraham]

I met the same problem such as @jolitte, when i try to decode it report the error which signs " ValueError: invalid literal for int() with base 2: 'TrueFalseTrueTrueTrueTrueTrueFalse' ". I work this project in Windows 10 which installed python 3.8 and pytorch 1.8.1 based on cuda 11.
Then I turn to use pytorch-cpu 1.0.0, it works but the speed of encoding and decoding is too low.

@DG-Abraham We don't support python 3.8 officially and the torch dependency is set to 1.0.0 as we solved the issue #42

Nvidia RTX30 device only support Cuda 11, and the oldest version of pytorch that support Cuda 11 is '1.7.0', so is there no way to solve this problem on RTX30 device? I try to decode on Cuda11, python 3.6, pytorch 1.7.0, it still not works with the same bug.

[Vipin0701]

Hi @pvk-developer, Could you please guide me through the configuration of Steganogan on local. As I am looking for generating the multiple stego images using steganogan. Thanks