ppforest2::cli::ConvergenceParams Struct Reference

Convergence and iteration control (evaluate + benchmark). More...

#include <EvaluateParams.hpp>

Public Attributes
float	cv = 0.05f
	CV target (e.g. 0.05 = stop when std < 5% of mean).
bool	enabled = true
	Adaptive stopping (default on; -i disables).
int	max = 200
	Hard upper bound on iterations.
int	min = 10
	Minimum iterations before checking convergence.
int	warmup = 0
	Warmup iterations discarded before measuring.
int	window = 3
	Consecutive checks below threshold before stopping.

Detailed Description

Convergence and iteration control (evaluate + benchmark).

See ConvergenceCriteria in Benchmark.hpp for the per-scenario equivalent used when parsing benchmark JSON files.

Member Data Documentation

cv

float ppforest2::cli::ConvergenceParams::cv = 0.05f

CV target (e.g. 0.05 = stop when std < 5% of mean).

enabled

bool ppforest2::cli::ConvergenceParams::enabled = true

Adaptive stopping (default on; -i disables).

max

int ppforest2::cli::ConvergenceParams::max = 200

Hard upper bound on iterations.

min

int ppforest2::cli::ConvergenceParams::min = 10

Minimum iterations before checking convergence.

warmup

int ppforest2::cli::ConvergenceParams::warmup = 0

Warmup iterations discarded before measuring.

window

int ppforest2::cli::ConvergenceParams::window = 3

Consecutive checks below threshold before stopping.

---

The documentation for this struct was generated from the following file:

• core/src/cli/EvaluateParams.hpp