reusex::vision::AnnotationConfig Struct Reference

Configuration for ML annotation inference and data loading. More...

#include <annotate.hpp>

Public Attributes
bool	use_cuda = false
	Force CUDA acceleration.
size_t	batch_size = 16
	Batch size for inference (recommended: 8-64).
bool	shuffle = false
	Shuffle dataset before processing.
size_t	num_workers = 4
	Number of worker threads (recommended: 2-4).
size_t	prefetch_batches = 8
	Batches to prefetch (recommended: 2-3x workers).

Detailed Description

Configuration for ML annotation inference and data loading.

Definition at line 11 of file annotate.hpp.

Member Data Documentation

batch_size

size_t reusex::vision::AnnotationConfig::batch_size = 16

Batch size for inference (recommended: 8-64).

Definition at line 13 of file annotate.hpp.

num_workers

size_t reusex::vision::AnnotationConfig::num_workers = 4

Number of worker threads (recommended: 2-4).

Definition at line 15 of file annotate.hpp.

prefetch_batches

size_t reusex::vision::AnnotationConfig::prefetch_batches = 8

Batches to prefetch (recommended: 2-3x workers).

Definition at line 16 of file annotate.hpp.

shuffle

bool reusex::vision::AnnotationConfig::shuffle = false

Shuffle dataset before processing.

Definition at line 14 of file annotate.hpp.

use_cuda

bool reusex::vision::AnnotationConfig::use_cuda = false

Force CUDA acceleration.

Definition at line 12 of file annotate.hpp.

---

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

• libs/reusex/include/vision/annotate.hpp