Number of calibration batches

Description

App Configuration Pane: Deep Learning

Configuration Objects: coder.TensorRTConfig

Numeric value specifying the number of batches for int8 calibration. The software uses the product of batchsize*NumCalibrationBatches to pick a random subset of images from the image dataset to perform calibration. The batchsize*NumCalibrationBatches value must not be greater than the number of images present in the image dataset. This option is applicable only when DataType is set to 'int8'.

NVIDIA^® recommends that about 500 images are sufficient for calibrating. Refer to the TensorRT documentation for more information.

Dependencies

To enable this parameter, you must set both these parameters:

Deep learning library to TensorRT
Data type to int8

Settings

50: This setting is the default setting.
positive integer: Numeric value specifying the number of batches for int8 calibration, specified as a positive integer.

Programmatic Use

Property: NumCalibrationBatches

Values: 50 | positive integer

Default: 50

Version History

Introduced in R2018b