R/serverless_model.R

In DyfanJones/sagemaker-r-mlcore: sagemaker machine learning core classes and methods

#' # NOTE: This code has been modified from AWS Sagemaker Python:
#' # https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/serverless/model.py
#'
#' #' @include r_utils.R
#' #' @include model.R
#' #' @include serverless_predictor.R
#'
#' #' @import R6
#' #' @import sagemaker.core
#'
#' #' @title LambdaModel class
#' #' @description A model that can be deployed to Lambda
#' #' @export
#' LambdaModel = R6Class("LambdaModel",
#'   inherit = ModelBase,
#'   public = list(
#'
#'     #' @description Initialize instance attributes.
#'     #' @param image_uri : URI of a container image in the Amazon ECR registry. The image
#'     #'              should contain a handler that performs inference.
#'     #' @param role : The Amazon Resource Name (ARN) of the IAM role that Lambda will assume
#'     #'              when it performs inference
#'     #' @param sagemaker_session : sagemaker session.
#'     initialize = function(image_uri,
#'                           role,
#'                           sagemaker_session=NULL){
#'       self$sagemaker_session = sagemaker_session %||% sagemaker.core::Session$new()
#'       self$client = self$sagemaker_session$paws_session$client("lambda")
#'       private$.image_uri = image_uri
#'       private$.role = role
#'     },
#'
#'     #' @description Create a Lambda function using the image specified in the constructor.
#'     #' @param function_name : The name of the function.
#'     #' @param timeout : The number of seconds that the function can run for before being terminated.
#'     #' @param memory_size : The amount of memory in MB that the function has access to.
#'     #' @param wait : If true, wait until the deployment completes (default: True).
#'     #' @return A LambdaPredictor instance that performs inference using the specified image.
#'     deploy = function(function_name,
#'                       timeout,
#'                       memory_size,
#'                       wait = TRUE){
#'       response = private$.client$create_function(
#'         FunctionName=function_name,
#'         PackageType="Image",
#'         Role=private$.role,
#'         Code=list(
#'           "ImageUri"=private$.image_uri),
#'         Timeout=timeout,
#'         MemorySize=memory_size
#'       )
#'
#'       if (!wait)
#'         return(LambdaPredictor$new(function_name, client=private$.client))
#'
#'       # Poll function state.
#'       polling_interval = 5
#'       while (response[["State"]] == "Pending"){
#'         Sys.sleep(polling_interval)
#'         response = private$.client$get_function_configuration(FunctionName=function_name)
#'       }
#'
#'       if (response[["State"]] != "Active")
#'         RuntimeError$new(sprintf("Failed to deploy model to Lambda: %s", response[["StateReason"]]))
#'
#'       return(LambdaPredictor$new(function_name, client=private$.client))
#'     },
#'
#'     #' @description Destroy resources associated with this model.
#'     #'              This method does not delete the image specified in the constructor. As
#'     #'              a result, this method is a no-op.
#'     delete_model = function(){
#'       return(invisible(NULL))
#'     }
#'   ),
#'
#'   private = list(
#'     .client = NULL,
#'     .image_uri = NULL,
#'     .role = NULL
#'   )
#' )