arnaghosh / auto-encoder Goto Github PK

Auto-encoder on torch - trying out the various AEs

Lua 65.87% MATLAB 0.44% Python 33.68%

auto-encoder's Issues

Why fully connection in Decoder?

Hi, I find that there are three linear and bn at the beginning of Decoder (196-200 in resnet.py). Why adding these layers? Much Thanks.

Is there a License for this Repository?

Thank you for this great repository. I am using the Resnet Autoencoder implementation code for my academic work, but I did not see any license or permission regarding academic usage. Could you please clarify this issue for me?

You could choose a license from the GitHub documentation page.
https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/customizing-your-repository/licensing-a-repository

If you want it short and simple, many open-source repositories are using the MIT license (just a suggestion). https://choosealicense.com/licenses/mit/

decoder input shape

When running resnet.py, there is following error:

Traceback (most recent call last):
  File "/home/wentian/.pycharm_helpers/pydev/pydevd.py", line 1741, in <module>
    main()
  File "/home/wentian/.pycharm_helpers/pydev/pydevd.py", line 1735, in main
    globals = debugger.run(setup['file'], None, None, is_module)
  File "/home/wentian/.pycharm_helpers/pydev/pydevd.py", line 1135, in run
    pydev_imports.execfile(file, globals, locals)  # execute the script
  File "/home/wentian/.pycharm_helpers/pydev/_pydev_imps/_pydev_execfile.py", line 18, in execfile
    exec(compile(contents+"\n", file, 'exec'), glob, loc)
  File "/home/wentian/work/caption_t_r_2/src/test.py", line 359, in <module>
    pred = autoencoder(inputs)
  File "/home/wentian/library/miniconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 491, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/wentian/work/caption_t_r_2/src/test.py", line 262, in forward
    x = self.decoder(x)
  File "/home/wentian/library/miniconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 491, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/wentian/work/caption_t_r_2/src/test.py", line 213, in forward
    x = F.relu(self.bn3(x))
  File "/home/wentian/library/miniconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 491, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/wentian/library/miniconda3/lib/python3.6/site-packages/torch/nn/modules/batchnorm.py", line 45, in forward
    self._check_input_dim(input)
  File "/home/wentian/library/miniconda3/lib/python3.6/site-packages/torch/nn/modules/batchnorm.py", line 193, in _check_input_dim
    .format(input.dim()))
ValueError: expected 4D input (got 2D input)

At line 262, the shape of x is (11, 48).

Binary function?

I tried to understand why did you put this Binary layer between encoder and decoder. I mean are you trying to create a map of positive values? If this is the case, I am not sure if restricting them will work better than just freeing them up, also wouldn't using something like x[x > 0] or something be simpler?

arnaghosh / auto-encoder Goto Github PK

auto-encoder's People

Contributors

Stargazers

Watchers

Forkers

auto-encoder's Issues

Why fully connection in Decoder?

Is there a License for this Repository?

decoder input shape

Binary function?

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent