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

Module: tfx.v1.proto

TFX proto module.

Classes

class ClassifyOutput: One type of output_type under proto.OutputColumnsSpec.

class CustomConfig: Optional specified configuration for ExampleGen components.

class DataSpec: Indicates which splits of examples should be processed incomponents.BulkInferrer.

class EnvVar: EnvVar represents an environment variable present in a Container.

class EnvVarSource: EnvVarSource represents a source for the value of an EnvVar.

class EvalArgs: Args specific to eval in components.Trainer.

class FeatureSlicingSpec: Slices corresponding to data set in components.Evaluator.

class Filesystem: File system based destination definition.

class Input: Specification of the input of the ExampleGen components.

class KubernetesConfig: Kubernetes configuration. We currently only support the use case when infra validator is run by orchestration.KubeflowDagRunner.

class LocalDockerConfig: Docker runtime in a local machine. This is useful when you're running pipeline with infra validator component in your your local machine.

class ModelSpec: Specifies the signature name to run the inference in components.BulkInferrer.

class Output: Specification of the output of the ExampleGen components.

class OutputColumnsSpec: The signature_name should exist in ModelSpec.model_signature_name.

class OutputExampleSpec: Defines how the inferrence results map to columns in output example in components.BulkInferrer.

class PodOverrides: Flattened collections of overridable variables for Pod and its submessages.

class PredictOutput: One type of output_type under proto.OutputColumnsSpec.

class PredictOutputCol: Proto type of output_columns under proto.PredictOutput.

class PushDestination: Defines the destination of pusher in components.Pusher.

class RangeConfig: RangeConfig is an abstract proto which can be used to describe ranges for different entities in TFX Pipeline.

class RegressOutput: One type of output_type under proto.OutputColumnsSpec.

class RequestSpec: Optional configuration about making requests from examples input in components.InfraValidator.

class RollingRange: Describes a rolling range.

class SecretKeySelector: SecretKeySelector selects a key of a Secret.

class ServingSpec: Defines an environment of the validating infrastructure in components.InfraValidator.

class SingleSlicingSpec: Specifies a single directive for choosing features for slicing.

class SplitConfig: A config to partition examples into split in proto.Output of ExampleGen.

class SplitsConfig: Defines the splits config in components.Transform.

class StaticRange: Describes a static window within the specified span numbers [start_span_number, end_span_number].

class TensorFlowServing: TensorFlow Serving docker image (tensorflow/serving) for serving binary.

class TensorFlowServingRequestSpec: Request spec for building TF Serving requests.

class TrainArgs: Args specific to training in components.Trainer.

class TuneArgs: Args specific to tuning in components.Tuner.

class ValidationSpec: Specification for validation criteria and thresholds in components.InfraValidator.

PayloadFormat Instance of google.protobuf.internal.enum_type_wrapper.EnumTypeWrapper

Enum to indicate payload format ExampleGen produces.

Versioning Instance of google.protobuf.internal.enum_type_wrapper.EnumTypeWrapper

Versioning method for the model to be pushed. Note that This is the semantic TFX provides, therefore depending on the platform, some versioning method might not be compatible. For example TF Serving only accepts an integer version that is monotonically increasing.