`neural_compressor.experimental.common.criterion`

Initialize critetion classes.

Classes includes:: TensorFlowCrossEntropyLoss, PyTorchCrossEntropyLoss, TensorflowKnowledgeDistillationLoss, PyTorchKnowledgeDistillationLoss, PyTorchIntermediateLayersKnowledgeDistillationLoss.

Module Contents

`TensorflowCriterions`	Record criterions in TensorflowCriterions class.
`PyTorchCriterions`	Record criterions in PyTorchCriterions class.
`Criterions`	Integrate criterions of different framework.
`KnowledgeDistillationFramework`	Knowledge Distillation Framework.
`KnowledgeDistillationLoss`	Initialize the KnowledgeDistillationLoss class.
`PyTorchKnowledgeDistillationLoss`	The PyTorchKnowledgeDistillationLoss class inherits from KnowledgeDistillationLoss.
`PyTorchKnowledgeDistillationLossWrapper`	PyTorchKnowledgeDistillationLossWrapper wraps PyTorchKnowledgeDistillationLoss.
`TensorflowKnowledgeDistillationLossExternal`	TensorflowKnowledgeDistillationLossExternal inherits from KnowledgeDistillationLoss.
`IntermediateLayersKnowledgeDistillationLoss`	The IntermediateLayersKnowledgeDistillationLoss class inherits from KnowledgeDistillationLoss.
`PyTorchIntermediateLayersKnowledgeDistillationLoss`	PyTorch Intermediate Layers Knowledge Distillation Loss.
`PyTorchIntermediateLayersKnowledgeDistillationLossWrapper`	PyTorch Intermediate Layers Knowledge Distillation Loss Wrapper.
`SelfKnowledgeDistillationLoss`	SelfKnowledge Distillation Loss.

criterion_registry(criterion_type, framework)

Use to register criterion classes in registry_criterions.

class neural_compressor.experimental.common.criterion.TensorflowCriterions[source]: Record criterions in TensorflowCriterions class.

class neural_compressor.experimental.common.criterion.PyTorchCriterions[source]: Record criterions in PyTorchCriterions class.

class neural_compressor.experimental.common.criterion.Criterions(framework)[source]: Integrate criterions of different framework.

neural_compressor.experimental.common.criterion.criterion_registry(criterion_type, framework)[source]

Use to register criterion classes in registry_criterions.

Parameters:

Returns:

The class of register.

Return type:

cls

class neural_compressor.experimental.common.criterion.KnowledgeDistillationFramework(student_model=None, teacher_model=None)[source]: Knowledge Distillation Framework.

class neural_compressor.experimental.common.criterion.KnowledgeDistillationLoss(temperature=1.0, loss_types=['CE', 'CE'], loss_weights=[0.5, 0.5], student_model=None, teacher_model=None)[source]: Initialize the KnowledgeDistillationLoss class.

class neural_compressor.experimental.common.criterion.PyTorchKnowledgeDistillationLoss(temperature=1.0, loss_types=['CE', 'CE'], loss_weights=[0.5, 0.5], student_model=None, teacher_model=None)[source]: The PyTorchKnowledgeDistillationLoss class inherits from KnowledgeDistillationLoss.

class neural_compressor.experimental.common.criterion.PyTorchKnowledgeDistillationLossWrapper(param_dict)[source]: PyTorchKnowledgeDistillationLossWrapper wraps PyTorchKnowledgeDistillationLoss.

class neural_compressor.experimental.common.criterion.TensorflowKnowledgeDistillationLossExternal(temperature=1.0, loss_types=['CE', 'CE'], loss_weights=[0.5, 0.5], student_model=None, teacher_model=None)[source]: TensorflowKnowledgeDistillationLossExternal inherits from KnowledgeDistillationLoss.

class neural_compressor.experimental.common.criterion.IntermediateLayersKnowledgeDistillationLoss(layer_mappings=[], loss_types=None, loss_weights=None, add_origin_loss=False, student_model=None, teacher_model=None)[source]: The IntermediateLayersKnowledgeDistillationLoss class inherits from KnowledgeDistillationLoss.

class neural_compressor.experimental.common.criterion.PyTorchIntermediateLayersKnowledgeDistillationLoss(layer_mappings=[], loss_types=None, loss_weights=None, add_origin_loss=False, student_model=None, teacher_model=None)[source]: PyTorch Intermediate Layers Knowledge Distillation Loss.

class neural_compressor.experimental.common.criterion.PyTorchIntermediateLayersKnowledgeDistillationLossWrapper(param_dict)[source]: PyTorch Intermediate Layers Knowledge Distillation Loss Wrapper.

class neural_compressor.experimental.common.criterion.SelfKnowledgeDistillationLoss(layer_mappings=[], loss_types=None, loss_weights=None, temperature=1.0, add_origin_loss=False, student_model=None, teacher_model=None)[source]: SelfKnowledge Distillation Loss.