Join the SIG TFX-Addons community and help make TFX even better!

tfx.components.Trainer.SPEC_CLASS

Trainer component spec.

Inherits From: ComponentSpec

tfx.components.Trainer.SPEC_CLASS(
    **kwargs
)

**kwargs Any inputs, outputs and execution parameters for this instance of the component spec.

Methods

from_json_dict

View source

@classmethod
from_json_dict(
    dict_data: Dict[Text, Any]
) -> Any

Convert from dictionary data to an object.

to_json_dict

View source

to_json_dict() -> Dict[Text, Any]

Convert from an object to a JSON serializable dictionary.

__eq__

View source

__eq__(
    other
)

Return self==value.

INPUTS 

{
 'base_model': ChannelParameter(type: <class 'tfx.types.standard_artifacts.Model'>),
 'examples': ChannelParameter(type: <class 'tfx.types.standard_artifacts.Examples'>),
 'hyperparameters': ChannelParameter(type: <class 'tfx.types.standard_artifacts.HyperParameters'>),
 'schema': ChannelParameter(type: <class 'tfx.types.standard_artifacts.Schema'>),
 'transform_graph': ChannelParameter(type: <class 'tfx.types.standard_artifacts.TransformGraph'>)
}

OUTPUTS 

{
 'model': ChannelParameter(type: <class 'tfx.types.standard_artifacts.Model'>),
 'model_run': ChannelParameter(type: <class 'tfx.types.standard_artifacts.ModelRun'>)
}

PARAMETERS 

{
 'custom_config': ExecutionParameter(type: (<class 'str'>, <class 'str'>), optional: True),
 'eval_args': ExecutionParameter(type: <class 'tfx.proto.trainer_pb2.EvalArgs'>, optional: False),
 'module_file': ExecutionParameter(type: (<class 'str'>, <class 'str'>), optional: True),
 'run_fn': ExecutionParameter(type: (<class 'str'>, <class 'str'>), optional: True),
 'train_args': ExecutionParameter(type: <class 'tfx.proto.trainer_pb2.TrainArgs'>, optional: False),
 'trainer_fn': ExecutionParameter(type: (<class 'str'>, <class 'str'>), optional: True)
}