slimgroup / InvertibleNetworks.jl

A Julia framework for invertible neural networks
MIT License
148 stars 20 forks source link

Glow 3d #61

Closed rafaelorozco closed 1 year ago

rafaelorozco commented 2 years ago

Small fixes to get glow working for 3d data. Should work for n-dim tensor as well but only tested with 3d.

codecov[bot] commented 2 years ago

Codecov Report

Merging #61 (dd14c31) into master (7070ec8) will increase coverage by 1.93%. The diff coverage is 91.69%.

@@            Coverage Diff             @@
##           master      #61      +/-   ##
==========================================
+ Coverage   85.94%   87.88%   +1.93%     
==========================================
  Files          31       31              
  Lines        2511     2327     -184     
==========================================
- Hits         2158     2045     -113     
+ Misses        353      282      -71     
Impacted Files Coverage Δ
src/InvertibleNetworks.jl 60.00% <0.00%> (ø)
src/conditional_layers/conditional_layer_hint.jl 99.18% <ø> (+4.33%) :arrow_up:
...itional_layers/conditional_layer_residual_block.jl 100.00% <ø> (+9.19%) :arrow_up:
src/layers/invertible_layer_basic.jl 93.33% <ø> (+1.12%) :arrow_up:
src/layers/invertible_layer_hint.jl 94.31% <ø> (+1.48%) :arrow_up:
src/layers/invertible_layer_hyperbolic.jl 86.20% <ø> (+2.04%) :arrow_up:
src/layers/invertible_layer_irim.jl 98.11% <ø> (+4.67%) :arrow_up:
src/layers/invertible_layer_template.jl 0.00% <ø> (ø)
src/layers/layer_affine.jl 97.22% <ø> (+7.22%) :arrow_up:
src/layers/layer_flux_block.jl 71.05% <ø> (ø)
... and 22 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

mloubout commented 1 year ago

Sweet thanks