Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

For a 0.5.0 release #259

Merged
merged 92 commits into from
Jun 11, 2024
Merged
Changes from 1 commit
Commits
Show all changes
92 commits
Select commit Hold shift + click to select a range
643c3a3
fix broken link in readme
ablaom Oct 31, 2023
c127b67
Merge branch 'dev' of https://github.com/FluxML/MLJFlux.jl into dev
ablaom Oct 31, 2023
a8b336e
add NeuralNetworkBinaryClassifier
tiemvanderdeure Apr 23, 2024
b8f0222
bump [compat] for Metahead
ablaom Apr 23, 2024
ac1c8b1
fix Metalhead breakage
ablaom Apr 23, 2024
b012344
add julia 1.9 to ci matrix
ablaom Apr 23, 2024
c1b0877
add [compat] for Statistics
ablaom Apr 23, 2024
33891e8
oops
ablaom Apr 23, 2024
b2bee39
dump nightly in testing
ablaom Apr 23, 2024
0a6e55d
Merge pull request #250 from FluxML/fix-metalhead-breakage
ablaom Apr 23, 2024
774c03d
Merge branch 'FluxML:dev' into binaryclassifier
tiemvanderdeure Apr 24, 2024
8ed2d15
add Optimisers to Project [deps]
ablaom Apr 29, 2024
8924a61
use explicit AD; refactor optimisers; disable l1/l2 reglrztn temporarily
ablaom Apr 30, 2024
52c0078
test adjustment: GPU's don't support StableRNGs
ablaom May 1, 2024
1ed8b49
further RNG fixes in tests
ablaom May 1, 2024
14615f9
remove invalid test of optimiser reproducibility on GPU
ablaom May 1, 2024
bf8f461
oops - forgotten `end`
ablaom May 1, 2024
bd5aa8b
oops again - use `accel` instead of `acceleration`
ablaom May 1, 2024
c806dc5
add some debugging lines
ablaom May 1, 2024
99100b8
use default_rng when comparing acceleration methods
ablaom May 1, 2024
f60be3f
disable test comparing losses b/w CPU/GPU in presence of dropout
ablaom May 1, 2024
3c39a26
oops. disabled test for wrong model
ablaom May 1, 2024
bf66132
fix test no longer broken; dump some redundant testing of optimiser
ablaom May 1, 2024
14b1993
revert to using StableRNGs when comparing accelerations for images
ablaom May 1, 2024
ffb20bd
refactor regularization (loss penalties)
ablaom May 3, 2024
b3b41ac
simplify: use MLP builder instead of custom Short2 in tests
ablaom May 3, 2024
103cbb4
🚀 Initialize documentation
EssamWisam May 22, 2024
359cb35
🎨 Add logo
EssamWisam May 22, 2024
56e65fb
🎨 Minor style improvements
EssamWisam May 22, 2024
3946a68
✨ Landing page and style changes
EssamWisam May 22, 2024
ee3e0b8
✨ Add basic docs skeleton and README integration
EssamWisam May 22, 2024
fd8ab78
📝 Add minor improvements
EssamWisam May 22, 2024
1dfad77
update examples/MNIST Manifest, including Julia 1.10
ablaom May 24, 2024
fccbfba
Update docs/src/index.md
EssamWisam May 24, 2024
ff5ee4e
Update docs/src/index.md
EssamWisam May 24, 2024
9a0d185
Update docs/src/index.md
EssamWisam May 24, 2024
111d507
Fix quick start by removing fit
EssamWisam May 24, 2024
3df8eb5
Merge branch 'instate-docs' of https://github.com/EssamWisam/MLJFlux.…
EssamWisam May 24, 2024
197dda8
Update docs/src/index.md
EssamWisam May 24, 2024
f2e0d91
Update docs/src/index.md
EssamWisam May 24, 2024
9fd8135
Update docs/src/interface/Summary.md
EssamWisam May 24, 2024
9332eaf
Update docs/src/interface/Custom Builders.md
EssamWisam May 24, 2024
c3b65e9
Update docs/src/interface/Custom Builders.md
EssamWisam May 24, 2024
4ad4ed5
Update docs/src/interface/Custom Builders.md
EssamWisam May 24, 2024
ca78165
Update docs/src/interface/Custom Builders.md
EssamWisam May 24, 2024
e68755c
Update docs/src/interface/Custom Builders.md
EssamWisam May 24, 2024
640ef2c
Update docs/src/workflow examples/Incremental Training.md
EssamWisam May 24, 2024
74e8284
Update docs/src/workflow examples/Incremental Training.md
EssamWisam May 24, 2024
5cc3dcd
Update docs/src/full tutorials/MNIST.md
EssamWisam May 24, 2024
211f505
Update docs/src/full tutorials/MNIST.md
EssamWisam May 24, 2024
2dd1c2a
Update docs/src/full tutorials/MNIST.md
EssamWisam May 24, 2024
8660d5b
Update docs/src/full tutorials/MNIST.md
EssamWisam May 24, 2024
560c748
Update docs/make.jl
EssamWisam May 24, 2024
2c8c2b9
🚑 Minor fixes
EssamWisam May 25, 2024
eb0f3e2
✨ Attempt deployment
EssamWisam May 25, 2024
ec6004f
🚑 Towards documentation
EssamWisam May 25, 2024
01ad08e
Merge pull request #252 from EssamWisam/instate-docs
EssamWisam May 25, 2024
638a19c
add some clarificatioins about GPU case in MNIST example
May 26, 2024
3f16411
fix a Literate issue with code comments
ablaom May 26, 2024
decef9b
Merge branch 'FluxML:dev' into binaryclassifier
tiemvanderdeure May 28, 2024
4af84e5
address review: add logic to regularized_optimiser()
ablaom May 29, 2024
943965d
update docstrings for `train` and `train_epoch`
ablaom May 29, 2024
5842ed9
add some docs for the binary classifier
tiemvanderdeure May 30, 2024
d5fe2c7
✨ Add 7 workflow examples for MLJFlux
EssamWisam Jun 3, 2024
975ed6e
✨ Add RNNs for Spam Classification Tutorial
EssamWisam Jun 4, 2024
4fcc61a
introduce variables for the test and train indices
ablaom Jun 6, 2024
9f76694
commit review suggestion
ablaom Jun 6, 2024
d3e4662
address remaining review comments. New notebook.ipynb not executed!
ablaom Jun 6, 2024
ed4b9bd
Merge branch 'mnist-update' of https://github.com/FluxML/MLJFlux.jl i…
ablaom Jun 6, 2024
0afbff1
generate a new notebook.ipynb (executed) for examples/mnist
ablaom Jun 6, 2024
e86009d
Update docs/src/full tutorials/Spam Detection with RNNs/SMS.jl
EssamWisam Jun 7, 2024
e1a2d6f
Update docs/src/full tutorials/Spam Detection with RNNs/SMS.jl
EssamWisam Jun 7, 2024
79f85d3
Update docs/src/workflow examples/Composition/composition.jl
EssamWisam Jun 7, 2024
e476125
Update docs/src/workflow examples/Early Stopping/iteration.jl
EssamWisam Jun 7, 2024
2698413
Update docs/src/workflow examples/Live Training/live-training.jl
EssamWisam Jun 7, 2024
958b9fe
✏️ Fix indentations
EssamWisam Jun 7, 2024
d737f71
🔥 Remove Literate dependency
EssamWisam Jun 7, 2024
801bc18
🚑 No need to load standardizer
EssamWisam Jun 7, 2024
1bd58dd
deprecating fit! and train! methods
pat-alt Jun 10, 2024
f38e0cf
Merge pull request #251 from FluxML/refactor-regularization
ablaom Jun 10, 2024
e182bd0
✏️ Attempt for a better link for Quick Start
EssamWisam Jun 10, 2024
47e52db
🎨 Improve styles for drop downs
EssamWisam Jun 10, 2024
38ae2d5
🚧 Hide spam classification tutorial temporarily
EssamWisam Jun 10, 2024
dd00443
Merge branch 'binaryclassifier' of https://github.com/tiemvanderdeure…
ablaom Jun 10, 2024
3ac787e
Merge pull request #256 from EssamWisam/more-tutorials
EssamWisam Jun 10, 2024
b7291f6
fix defaults for NNBinaryClassifier constructor
ablaom Jun 10, 2024
7eae840
tweak NNBinaryClassifier to address breakages of #251
ablaom Jun 10, 2024
05be138
Merge pull request #257 from FluxML/tiemvanderdeure-binaryclassifier
ablaom Jun 10, 2024
088e152
Merge pull request #254 from FluxML/mnist-update
ablaom Jun 10, 2024
0b8be2a
bump 0.5.0
ablaom Jun 10, 2024
1d2670a
add check that Flux optimiser is not being used
ablaom Jun 11, 2024
06282cd
Merge pull request #260 from FluxML/optimiser-warning
ablaom Jun 11, 2024
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
disable test comparing losses b/w CPU/GPU in presence of dropout
  • Loading branch information
ablaom committed May 1, 2024
commit f60be3ff3800d1eb386736a5407d278d9766a628
4 changes: 2 additions & 2 deletions test/image.jl
Original file line number Diff line number Diff line change
Expand Up @@ -74,7 +74,7 @@ end
# check different resources (CPU1, CUDALibs) give about the same loss:
reference = losses[1]
@info "Losses for each computational resource: $losses"
@test all(x->abs(x - reference)/reference < 5e-4, losses[2:end])
# @test_broken all(x->abs(x - reference)/reference < 5e-4, losses[2:end])


# # BASIC IMAGE TESTS COLOR
Expand Down Expand Up @@ -118,7 +118,7 @@ end
# check different resources (CPU1, CUDALibs, etc)) give about the same loss:
reference = losses[1]
@info "Losses for each computational resource: $losses"
@test all(x->abs(x - reference)/reference < 1e-5, losses[2:end])
# @test_broken all(x->abs(x - reference)/reference < 1e-5, losses[2:end])


# # SMOKE TEST FOR DEFAULT BUILDER
Expand Down
Loading