|View source on GitHub|
TFX Pusher executor to push the new TF model to a filesystem target.
tfx.components.pusher.executor.Executor( context=None )
The Pusher component is used to deploy a validated model to a filesystem
target or serving environment using tf.serving. Pusher depends on the outputs
of ModelValidator to determine if a model is ready to push. A model is
considered to be safe to push only if ModelValidator has marked it as BLESSED.
A push action delivers the model exports produced by Trainer to the
destination defined in the
push_destination of the component config.
To include Pusher in a TFX pipeline, configure your pipeline similar to https://github.com/tensorflow/tfx/blob/master/tfx/examples/chicago_taxi_pipeline/taxi_pipeline_simple.py#L104.
For more details on tf.serving itself, please refer to https://tensorflow.org/tfx/guide/pusher. For a tutuorial on TF Serving, please refer to https://www.tensorflow.org/tfx/guide/serving.
CheckBlessing( input_dict )
Check that model is blessed by upstream validators.
input_dict: Input dict from input key to a list of artifacts:
- model_blessing: A
ModelBlessingartifact from model validator or evaluator. Pusher looks for a custom property
blessedin the artifact to check it is safe to push.
- infra_blessing: An
InfraBlessingartifact from infra validator. Pusher looks for a custom proeprty
blessedin the artifact to determine whether the model is mechanically servable from the model server to which Pusher is going to push.
- model_blessing: A
True if the model is blessed by validator.
Do( input_dict, output_dict, exec_properties )
Push model to target directory if blessed.
input_dict: Input dict from input key to a list of artifacts, including:
- model_export: exported model from trainer.
- model_blessing: model blessing path from model_validator. A push action delivers the model exports produced by Trainer to the destination defined in component config.
output_dict: Output dict from key to a list of artifacts, including:
- model_push: A list of 'ModelPushPath' artifact of size one. It will include the model in this push execution if the model was pushed.
exec_properties: A dict of execution properties, including:
- push_destination: JSON string of pusher_pb2.PushDestination instance, providing instruction of destination to push model.