ReferenceΒΆ

This is the complete list of currently supported network operations in RocketML:

  • Core layers:
    • Input
    • Dense
  • Convolution layers:
    • Conv1D
    • Conv2D
    • Conv3D
    • DepthwiseConv1D
    • DepthwiseConv2D
    • DepthwiseConv3D
    • SeparableConv1D
    • SeparableConv2D
    • Conv1DTranspose
    • Conv2DTranspose
    • Conv3DTranspose
  • Pooling layers:
    • MaxPooling1D
    • MaxPooling2D
    • MaxPooling3D
    • AveragePooling1D
    • AveragePooling2D
    • AveragePooling3D
    • GlobalMaxPooling1D
    • GlobalMaxPooling2D
    • GlobalMaxPooling3D
    • GlobalAveragePooling1D
    • GlobalAveragePooling2D
    • GlobalAveragePooling3D
  • Normalization layers:
    • BatchNormalization
    • InstanceNormalization
    • LayerNormalization
  • Regularization layers:
    • Dropout
    • SpatialDropout2D
    • SpatialDropout3D
  • Reshaping layers:
    • Reshape
    • Flatten
    • UpSampling1D
    • UpSampling2D
    • UpSampling3D
    • ZeroPadding1D
    • ZeroPadding2D
    • ZeroPadding3D
    • Cropping1D
    • Cropping2D
    • Cropping3D
  • Merging layers
    • Add
    • Subtract
    • Multiply
    • Average
    • Concatenate
  • Activation layers:
    • ReLU
    • LeakyReLU
    • Softmax
    • Log_Softmax
    • Sigmoid
    • Tanh
    • Swish