Closed tsaycal closed 2 years ago
Merging #39 (4bbb84a) into main (b0ccafd) will increase coverage by
1.20%
. The diff coverage isn/a
.:exclamation: Current head 4bbb84a differs from pull request most recent head 8c3f390. Consider uploading reports for the commit 8c3f390 to get more accurate results
@@ Coverage Diff @@
## main #39 +/- ##
==========================================
+ Coverage 88.79% 90.00% +1.20%
==========================================
Files 23 23
Lines 1071 1050 -21
Branches 160 153 -7
==========================================
- Hits 951 945 -6
+ Misses 98 86 -12
+ Partials 22 19 -3
Impacted Files | Coverage Δ | |
---|---|---|
src/omlt/neuralnet/layers/full_space.py | 58.82% <0.00%> (-1.79%) |
:arrow_down: |
src/omlt/gbt/gbt_formulation.py | 93.60% <0.00%> (+9.92%) |
:arrow_up: |
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 b0ccafd...8c3f390. Read the comment docs.
@tsaycal we just merged in PR #33. Can you update NeuralNetworkFormulation
to FullSpaceNNFormulation
? If you don't need to update logsoftmax, you can also just use ReluBigMFormulation
for relu networks.
Also, is this PR only for mnist_example_convolutional.ipynb
and mnist_example_dense.ipynb
? If so, we should un-commit the other notebooks.
Models are built using NeuralNetworkFormulation.