neural_compressor.data.transforms

Neural Compressor Built-in transforms for multiple framework backends.

Submodules

• neural_compressor.data.transforms.imagenet_transform
• neural_compressor.data.transforms.postprocess
• neural_compressor.data.transforms.tokenization
• neural_compressor.data.transforms.transform

Package Contents

Classes

TRANSFORMS	Transforms collection class.
BaseTransform	The base class for transform.
Postprocess	Just collect the infos to construct a Postprocess.

Functions

transform_registry(transform_type, process, framework)

Class decorator used to register all transform subclasses.

class neural_compressor.data.transforms.TRANSFORMS(framework, process)

Bases: object

Transforms collection class.

Provide register method to register new Transforms and provide __getitem__ method to get Transforms according to Transforms type.

register(name, transform_cls)

Register new Transform according to Transforms type.

Parameters:

• name (str) – process name

• transform_cls (class) – process function wrapper class

class neural_compressor.data.transforms.BaseTransform

Bases: object

The base class for transform.

neural_compressor.data.transforms.transform_registry(transform_type, process, framework)

Class decorator used to register all transform subclasses.

Parameters:

• transform_type (str) – Transform registration name

• process (str) – support 3 process including ‘preprocess’, ‘postprocess’, ‘general’

• framework (str) – support 4 framework including ‘tensorflow’, ‘pytorch’, ‘mxnet’, ‘onnxrt’

• cls (class) – The class of register.

Returns:

The class of register.

Return type:

cls

class neural_compressor.data.transforms.Postprocess(postprocess_cls, name='user_postprocess', **kwargs)

Bases: object

Just collect the infos to construct a Postprocess.