nvflare.app_common.decomposers.common_decomposers module¶

Decomposers for types from app_common and Machine Learning libraries.

class Float32ScalarDecomposer[source]¶

Bases: NumpyScalarDecomposer

supported_type()[source]¶

Returns the type/class supported by this decomposer.

Returns:: The class (not instance) of supported type

class Float64ScalarDecomposer[source]¶

Bases: NumpyScalarDecomposer

supported_type()[source]¶

Returns the type/class supported by this decomposer.

Returns:: The class (not instance) of supported type

class Int32ScalarDecomposer[source]¶

Bases: NumpyScalarDecomposer

supported_type()[source]¶

Returns the type/class supported by this decomposer.

Returns:: The class (not instance) of supported type

class Int64ScalarDecomposer[source]¶

Bases: NumpyScalarDecomposer

supported_type()[source]¶

Returns the type/class supported by this decomposer.

Returns:: The class (not instance) of supported type

class ModelLearnableDecomposer[source]¶

Bases: Decomposer

decompose(target: ModelLearnable) → Any[source]¶

Decompose the target into types supported by msgpack or classes with decomposers registered.

Msgpack supports primitives, bytes, memoryview, lists, dicts.

Parameters:: target – The instance to be serialized
Returns:: The decomposed serializable objects

recompose(data: Any) → ModelLearnable[source]¶

Reconstruct the object from decomposed components.

Parameters:: data – The decomposed components
Returns:: The reconstructed object

supported_type()[source]¶

Returns the type/class supported by this decomposer.

Returns:: The class (not instance) of supported type

class NumpyArrayDecomposer[source]¶

Bases: Decomposer

decompose(target: ndarray) → Any[source]¶

Decompose the target into types supported by msgpack or classes with decomposers registered.

Msgpack supports primitives, bytes, memoryview, lists, dicts.

Parameters:: target – The instance to be serialized
Returns:: The decomposed serializable objects

recompose(data: Any) → ndarray[source]¶

Reconstruct the object from decomposed components.

Parameters:: data – The decomposed components
Returns:: The reconstructed object

supported_type()[source]¶

Returns the type/class supported by this decomposer.

Returns:: The class (not instance) of supported type

class NumpyScalarDecomposer[source]¶

Bases: Decomposer, ABC

Decomposer base class for all numpy types with item method.

decompose(target: Any) → Any[source]¶

Decompose the target into types supported by msgpack or classes with decomposers registered.

Msgpack supports primitives, bytes, memoryview, lists, dicts.

Parameters:: target – The instance to be serialized
Returns:: The decomposed serializable objects

recompose(data: Any) → ndarray[source]¶

Reconstruct the object from decomposed components.

Parameters:: data – The decomposed components
Returns:: The reconstructed object

register()[source]¶