NVIDIA Developer Forums

Calibrate INT8 with different inputs of different sizes

AI & Data Science Deep Learning (Training & Inference) TensorRT

hl2997 June 26, 2019, 2:04am 1

Hi, I am implementing a calibration plugin layer for calibrating Caffe model with int8.
Since the input size of the model is static after parsed, is there any way I can feed the input of different size in each batch?

Topic		Replies	Views	Activity
TensorRT 5 batch size TensorRT	1	971	March 25, 2019
Does TensorRT Python API support uff model int8 calibration TensorRT	1	976	July 11, 2019
Does TensorRt support flexible sizes of input ？ TensorRT	2	559	October 12, 2021
Examples of more diverse usage of TensorRT IInt8Calibrator available elsewhere? GPU-Accelerated Libraries	0	493	July 21, 2017
INT8 Calibration Cache format - could it be officially documented? TensorRT	5	627	August 1, 2023
int8 precision TensorRT	0	353	December 5, 2019
How to dynamic setup input size for a Caffe model? TensorRT	1	1209	October 10, 2018
How to create batch files for int 8 faster rcnn inference? TensorRT	0	452	June 4, 2018
I want to inference caffe model with INT8 precision, what should I do Jetson Xavier NX tensorrt	3	462	October 18, 2021
How do I generate INT8 calibration file wiht caffe? TensorRT tensorrt	1	830	August 12, 2020