merge_results: Combine the results from the parallel training.

Description Usage Arguments

View source: R/do_azureml_parallel.R

Description

Combine the results from the parallel training.

Usage

1
merge_results(node_count, process_count_per_node, run, source_directory)

Arguments

node_count

Number of nodes in the AmlCompute cluster.

process_count_per_node

Number of processes per node.

run

The run object whose output needs to be combined.

source_directory

The directory where the output from the run would be downloaded.


azuremlsdk documentation built on Oct. 23, 2020, 8:22 p.m.