| Submodule | Maintainers | Contact Info |
|---|---|---|
| gelu | @AakashKumarNain | aakashnain@outlook.com |
| maxout | @fsx950223 | fsx950223@gmail.com |
| normalizations | @smokrow | moritz.kroeger@tu-dortmund.de |
| opticalflow | @fsx950223 | fsx950223@gmail.com |
| poincare | @rahulunair | rahulunair@gmail.com |
| polynomial | @tanzheny | tanzheny@google.com |
| sparsemax | @AndreasMadsen | amwwebdk+github@gmail.com |
| tlu | @AakashKumarNain | aakashnain@outlook.com |
| wrappers | @seanpmorgan | seanmorgan@outlook.com |
| Submodule | Layer | Reference |
|---|---|---|
| gelu | GELU | https://arxiv.org/abs/1606.08415 |
| maxout | Maxout | https://arxiv.org/abs/1302.4389 |
| normalizations | GroupNormalization | https://arxiv.org/abs/1803.08494 |
| normalizations | InstanceNormalization | https://arxiv.org/abs/1607.08022 |
| opticalflow | CorrelationCost | https://arxiv.org/abs/1504.06852 |
| poincare | PoincareNormalize | https://arxiv.org/abs/1705.08039 |
| polynomial | PolynomialCrossing | https://arxiv.org/pdf/1708.05123 |
| sparsemax | Sparsemax | https://arxiv.org/abs/1602.02068 |
| tlu | TLU | https://arxiv.org/abs/1911.09737 |
| wrappers | WeightNormalization | https://arxiv.org/abs/1602.07868 |
In order to conform with the current API standard, all layers must:
- Inherit from either
keras.layers.Layeror its subclasses. - Register as a keras global object so it can be serialized properly:
@tf.keras.utils.register_keras_serializable(package='Addons') - Add the addon to the
py_libraryin this sub-package's BUILD file.
- Simple unittests that demonstrate the layer is behaving as expected.
- When applicable, run all unittests with TensorFlow's
@run_in_graph_and_eager_modes(for test method) orrun_all_in_graph_and_eager_modes(for TestCase subclass) decorator. - Run
layer_teston the layer. - Add a
py_testto this sub-package's BUILD file.
- Update the table of contents in this sub-package's README.