► Keras 2 API 文档 / 层 API
层 API
基础层类
- 层类
- weights 属性
- trainable_weights 属性
- non_trainable_weights 属性
- add_weight 方法
- trainable 属性
- get_weights 方法
- set_weights 方法
- get_config 方法
- add_loss 方法
- losses 属性
层激活函数
层权重初始化器
- RandomNormal 类
- RandomUniform 类
- TruncatedNormal 类
- Zeros 类
- Ones 类
- GlorotNormal 类
- GlorotUniform 类
- HeNormal 类
- HeUniform 类
- Identity 类
- Orthogonal 类
- Constant 类
- VarianceScaling 类
层权重正则化器
层权重约束
核心层
卷积层
- Conv1D 层
- Conv2D 层
- Conv3D 层
- SeparableConv1D 层
- SeparableConv2D 层
- DepthwiseConv2D 层
- Conv1DTranspose 层
- Conv2DTranspose 层
- Conv3DTranspose 层
池化层
- MaxPooling1D 层
- MaxPooling2D 层
- MaxPooling3D 层
- AveragePooling1D 层
- AveragePooling2D 层
- AveragePooling3D 层
- GlobalMaxPooling1D 层
- GlobalMaxPooling2D 层
- GlobalMaxPooling3D 层
- GlobalAveragePooling1D 层
- GlobalAveragePooling2D 层
- GlobalAveragePooling3D 层
循环层
预处理层
标准化层
正则化层
- Dropout 层
- SpatialDropout1D 层
- SpatialDropout2D 层
- SpatialDropout3D 层
- GaussianDropout 层
- GaussianNoise 层
- ActivityRegularization 层
注意力层
重塑层
- Reshape 层
- Flatten 层
- RepeatVector 层
- Permute 层
- Cropping1D 层
- Cropping2D 层
- Cropping3D 层
- UpSampling1D 层
- UpSampling2D 层
- UpSampling3D 层
- ZeroPadding1D 层
- ZeroPadding2D 层
- ZeroPadding3D 层