Skip to content

Conversation

@vitskvara
Copy link
Collaborator

No description provided.

@codecov-io
Copy link

codecov-io commented Sep 18, 2019

Codecov Report

Merging #31 into master will increase coverage by 3.68%.
The diff coverage is 65.71%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master      #31      +/-   ##
==========================================
+ Coverage   80.85%   84.53%   +3.68%     
==========================================
  Files          11       11              
  Lines         188      194       +6     
==========================================
+ Hits          152      164      +12     
+ Misses         36       30       -6
Impacted Files Coverage Δ
src/GenerativeModels.jl 100% <ø> (ø) ⬆️
src/models/vae.jl 94.73% <0%> (ø) ⬆️
src/pdfs/cgaussian.jl 100% <100%> (ø) ⬆️
src/pdfs/abstract_pdfs.jl 100% <100%> (ø) ⬆️
src/pdfs/gaussian.jl 100% <100%> (ø) ⬆️
src/pdfs/svar_cgaussian.jl 100% <100%> (ø) ⬆️
src/models/gan.jl 52.63% <52.63%> (ø)
src/utils/utils.jl 57.89% <66.66%> (+0.31%) ⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 83b6532...c0e18e4. Read the comment docs.

@nmheim
Copy link
Member

nmheim commented Sep 18, 2019

Thanks for the GAN and the batchsize fixes!
I needed 1c48b2c for encoders that always return batchsize=1 even if input is a 1d vector like this one:

    Chain(                                                                                                                                                                                                                                    
        x -> reshape(x, size(x, 1), 1, 1, size(x, 2)),                                        
        Conv((1, 1), 1=>16, act),                                                                                                                                                                                                             
        x -> cat(                                                                                                                                                                                                                             
            mean(x, dims=1),                                                                                                                                                                                                                  
            maximum(x, dims=1), dims=3),                                                                                                                                                                                                      
        x -> dropdims(x, dims=(1,2)),               # this always keeps the batch dimension                                                                                                                                                                                          
        Dense(2*16, zlen)                                                                                                                                                                                                                     
    )

hope thats fine! :)

@vitskvara vitskvara merged commit 426efc6 into master Sep 19, 2019
@vitskvara vitskvara deleted the gan branch September 19, 2019 07:37
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants