• o7_309257
  • 230.7KB
  • zip
  • 0
  • VIP专享
  • 0
  • 2022-04-01 11:11
keras-contrib:Keras社区贡献 Keras-contrib已过时。 使用 Keras-contrib的未来: 我们正在迁移到 。 请在查看公告。 该库是python深度学习库的官方扩展库。 它包含其他层,激活,损失函数,优化器等,这些在Keras本身尚不可用。 所有这些附加模块都可以与Keras核心模型和模块结合使用。 随着Keras-Contrib中社区贡献的经过测试,使用,验证以及其效用得到证明,它们可以集成到Keras核心存储库中。 为了保持Keras简洁,干净且功能强大,只有最有用的功能才能使它进入Keras。 这个贡献库既是新功能的试验场,也是功能(虽然有用)可
# keras-contrib : Keras community contributions Keras-contrib is deprecated. Use [TensorFlow Addons]( ## The future of Keras-contrib: We're migrating to [tensorflow/addons]( See the announcement [here]( [![Build Status](]( This library is the official extension repository for the python deep learning library [Keras]( It contains additional layers, activations, loss functions, optimizers, etc. which are not yet available within Keras itself. All of these additional modules can be used in conjunction with core Keras models and modules. As the community contributions in Keras-Contrib are tested, used, validated, and their utility proven, they may be integrated into the Keras core repository. In the interest of keeping Keras succinct, clean, and powerfully simple, only the most useful contributions make it into Keras. This contribution repository is both the proving ground for new functionality, and the archive for functionality that (while useful) may not fit well into the Keras paradigm. --- ## Installation #### Install keras_contrib for keras-team/keras For instructions on how to install Keras, see [the Keras installation page]( ```shell git clone cd keras-contrib python install ``` Alternatively, using pip: ```shell sudo pip install git+ ``` to uninstall: ```pip pip uninstall keras_contrib ``` #### Install keras_contrib for tensorflow.keras ```shell git clone cd keras-contrib python USE_TF_KERAS=1 python install ``` to uninstall: ```shell pip uninstall tf_keras_contrib ``` For contributor guidelines see []( --- ## Example Usage Modules from the Keras-Contrib library are used in the same way as modules within Keras itself. ```python from keras.models import Sequential from keras.layers import Dense import numpy as np # I wish Keras had the Parametric Exponential Linear activation.. # Oh, wait..! from keras_contrib.layers.advanced_activations import PELU # Create the Keras model, including the PELU advanced activation model = Sequential() model.add(Dense(100, input_shape=(10,))) model.add(PELU()) # Compile and fit on random data model.compile(loss='mse', optimizer='adam'), 10)), y=np.random.random((100, 100)), epochs=5, verbose=0) # Save our model'example.h5') ``` ### A Common "Gotcha" As Keras-Contrib is external to the Keras core, loading a model requires a bit more work. While a pure Keras model is loadable with nothing more than an import of `keras.models.load_model`, a model which contains a contributed module requires an additional import of `keras_contrib`: ```python # Required, as usual from keras.models import load_model # Recommended method; requires knowledge of the underlying architecture of the model from keras_contrib.layers import PELU from keras_contrib.layers import GroupNormalization # Load our model custom_objects = {'PELU': PELU, 'GroupNormalization': GroupNormalization} model = load_model('example.h5', custom_objects) ```