Monty.Experiment.RecognitionStrategy.RecognitionResult (ExTbpMonty.Core v0.1.0)

Copy Markdown View Source

What a policy decided: stop, time out the laggards, or start exploring.

Mirrors tbp/monty/experiment/recognition_strategy.py::RecognitionResult. Three booleans, and the episode loop does three different things with them:

  • is_done — stop the episode.
  • time_out_lms — tell the learning modules that have not concluded to conclude (:time_out) before the episode ends, so the episode still has an answer to report.
  • start_exploring — switch the model into exploring mode, which is how a training episode keeps moving over an object it has already recognised.

Summary

Types

t()

Types

t()

@type t() :: %Monty.Experiment.RecognitionStrategy.RecognitionResult{
  is_done: boolean(),
  start_exploring: boolean(),
  time_out_lms: boolean()
}