gdServiceTrain: Send a request to software service for accelerated training...

In ganGenerativeData: Generate Generative Data for a Data Source

Send a request to software service for accelerated training of generative models to train a generative model for a data source

Description

Send a request to software service to train a generative model. A data source file will be uploaded and a job for training will be generated. A job id for the generated job will be returned which has to be used in related requests. The job will be processed as soon as other waiting jobs are processed. When a name of an existing generative model file is passed the file will be also uploaded and the job will continue the training. See also functions gdTrain(), gdServiceGetGenerativeModel(), gdServiceGetGenerativeData(), gdServiceGetStatus(), gdServiceDelete().

Usage

gdServiceTrain(
  url,
  accessKey,
  generativeModelFileName,
  generativeDataFileName,
  dataSourceFileName,
  trainParameters = gdTrainParameters(numberOfTrainingIterations = 10000,
    numberOfInitializationIterations = 2500, numberOfHiddenLayerUnits = 1024,
    learningRate = 7e-05, dropout = 0.05)
)

Arguments

url	URL of software service for accelerated training of generative models
accessKey	Unique key for access to software service
generativeModelFileName	Name of generative model file
generativeDataFileName	Name of generative data file. If name is NULL or empty string generated data will not be written to a file.
dataSourceFileName	Name of data source file
trainParameters	Generative model training parameters, see function gdTrainParameters().

Value

Job Id number

Examples

## Not run: 
url <- "http://xxx.xxx.xxx.xxx/gdService"
accessKey <- "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
trainParameters <- gdTrainParameters(numberOfTrainingIterations = 10000)
jobId <- gdServiceTrain(url, accessKey, "gm.bin", "gd.bin", "ds.bin", trainParameters)
## End(Not run)

ganGenerativeData documentation built on Oct. 7, 2024, 5:09 p.m.