GPUMemoryIncrease: Debugger GPUMemoryIncrease class

In DyfanJones/sagemaker-r-debugger: R6sagemaker debugger for sagemaker operations

Description

This rule helps to detect large increase in memory usage on GPUs. The rule computes the moving average of continous datapoints and compares it against the moving average of previous iteration.

Super class

sagemaker.debugger::ProfilerRuleBase -> GPUMemoryIncrease

Methods

Public methods

• GPUMemoryIncrease$new()

• GPUMemoryIncrease$clone()

Inherited methods

• sagemaker.debugger::ProfilerRuleBase$format()

---

Method new()

Initialize GPUMemoryIncrease class

Usage

GPUMemoryIncrease$new(
  increase = 5,
  patience = 1000,
  window = 10,
  scan_interval_us = 60 * 1000 * 1000
)

Arguments

increase

: defines the threshold for absolute memory increase.Default is 5%. So if moving average increase from 5% to 6%, the rule will fire.

patience

: defines how many continous datapoints to capture before Rule runs the first evluation. Default is 1000

window

: window size for computing moving average of continous datapoints

scan_interval_us

: interval with which timeline files are scanned. Default is 60000000 us.

---

Method clone()

The objects of this class are cloneable with this method.

Usage

GPUMemoryIncrease$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

DyfanJones/sagemaker-r-debugger documentation built on Jan. 20, 2022, 5:49 p.m.