sagemaker_hyperparameter_tuner: Start a Sagemaker Hyperparamter Tuning Job

In tmastny/sagemaker: R Interface for the AWS Sagemaker API

Description

Interface to sagemaker$tuner$HyperparameterTuner.

Usage

sagemaker_hyperparameter_tuner(
  estimator,
  split,
  hyperparameter_ranges,
  strategy = "Random",
  max_jobs = 10L,
  max_parallel_jobs = 2L,
  ...
)

Arguments

estimator	Sagemaker estimator from sagemaker_estimator.
split	Train/validation dataset split from s3_split.
hyperparameter_ranges	A named list of model hyperparameters with sagemaker_ranges for tuning.
strategy	Tuning strategy: "Random" or "Bayesian".
max_jobs	Number of unique models to train during tuning.
max_parallel_jobs	Number of models to train simultaneously.
...	Additional named arguments sent to the underlying API.

tmastny/sagemaker documentation built on July 15, 2020, 4:17 p.m.