package yggdrasil_decision_forests.model.generic_worker.proto

Get desktop application:
View/edit binary Protocol Buffers messages

• optional string request_id = 1

  Free identifier.

• oneof type

  • Request.TrainModel train_model = 2

    Train a model.

  • Request.EvaluateModel evaluate_model = 3

    Evaluate a model.

Used in: Request

• optional string dataset_path = 1

  Follow the signature of cli/evaluate.cc.

• optional string model_path = 2

• optional metric.proto.EvaluationOptions options = 3

Used in: Request

• optional proto.TrainingConfig train_config = 1

  Follow the signature of cli/train.cc.

• optional proto.DeploymentConfig deployment_config = 2

• optional string dataset_path = 3

• optional dataset.proto.DataSpecification dataspec = 4

• optional string model_base_path = 5

  Base directory for the output model. The effective directory is returned in the answer Result::TrainModel.

• optional string log_directory = 6

• optional proto.GenericHyperParameters generic_hyper_parameters = 7

• optional string valid_dataset_path = 8

• optional bool return_model_validation = 9

  If true, populate the "validation_evaluation" field with the model self-evaluation. This is equivalent (but more efficient), that loading the entire model and extracting the self evaluation.

• optional string request_id = 1

• oneof type

  • Result.TrainModel train_model = 2

  • Result.EvaluateModel evaluate_model = 3

Used in: Result

• optional metric.proto.EvaluationResults evaluation = 1

Used in: Result

• optional string model_path = 1

• optional metric.proto.EvaluationResults validation_evaluation = 2

• optional string temporary_directory = 1