Skip to content

Environment

The environment configuration contains values that determine the training environment, such as the number of workers for multi-processing and the GPU ids to be used. The following yaml is the environment configuration example.

environment: 
  seed: 1
  batch_size: 8
  num_workers: 4 
  gpus: 0, 1, 2, 3

Field list

Field Description
environment.seed (int) Random seed.
environment.batch_size (int) The number of samples in single batch input.
environment.num_workers (int) The number of multi-processing workers to be used by the data loader.
environment.gpus (str) GPU ids to use, this should be separated by commas.