Closed DrChainsaw closed 2 years ago
Merging #80 (451207d) into master (66f2fd9) will increase coverage by
0.40%
. The diff coverage is94.90%
.
@@ Coverage Diff @@
## master #80 +/- ##
==========================================
+ Coverage 94.78% 95.18% +0.40%
==========================================
Files 15 23 +8
Lines 1667 1931 +264
==========================================
+ Hits 1580 1838 +258
- Misses 87 93 +6
Impacted Files | Coverage Δ | |
---|---|---|
src/crossover/graph.jl | 100.00% <ø> (ø) |
|
src/evolve.jl | 100.00% <ø> (ø) |
|
src/iterators.jl | 93.58% <91.17%> (+9.62%) |
:arrow_up: |
src/batchsize.jl | 92.53% <92.53%> (ø) |
|
src/crossover/optimizer.jl | 92.85% <92.85%> (ø) |
|
src/shape.jl | 95.89% <93.61%> (-1.92%) |
:arrow_down: |
src/fitness.jl | 92.02% <93.75%> (-0.16%) |
:arrow_down: |
src/mutation/generic.jl | 93.93% <93.93%> (ø) |
|
src/visualize/callbacks.jl | 96.72% <94.28%> (-1.04%) |
:arrow_down: |
src/mutation/optimizer.jl | 94.73% <94.73%> (ø) |
|
... and 15 more |
:mega: Codecov can now indicate which changes are the most critical in Pull Requests. Learn more
Don't forget to remove Cbc depencency before merging.
Some utilities for having batch size as a hyperparameter as well as (trying) to make sure to select batch sizes which won't make training or validation go oom.
Breaking due to increased Julia compat bound. Previous version was anyways most likely not compatible with Julia 1.0.0.