YoloV4 int8 conversion issue

Timorl · January 11, 2022, 12:57pm

Description

I’m tring to convert YoloV4 model to int8,
When I’m converting onnx to a fp32/fp16 engine I get bit-exec resault.
When I’m tring to convert to int8 I get really bad results.

Environment

TensorRT Version: 7.2.3.4
GPU Type: 3080
Nvidia Driver Version:
CUDA Version: 11.4
CUDNN Version: 8.1.1
Operating System + Version: ubuntu 20.04
Python Version (if applicable): 3.8
TensorFlow Version (if applicable):
PyTorch Version (if applicable): 1.9
Baremetal or Container (if container which image + tag):

This is my calibration object.
calibrator.py (3.8 KB)

NVES · January 11, 2022, 1:38pm

Hi, Please refer to the below links to perform inference in INT8

Thanks!

Topic		Replies	Views
How to convert pytorch (YoloV5) model to INT8 TensorRT	5	4812	December 28, 2020
I would like to ask how to get a int8 model? Jetson Xavier NX tensorrt	9	718	September 27, 2021
The inference speed of yolov5 tensorrt has little difference between int8 and fp16 TensorRT tensorrt , cuda	1	1511	September 8, 2022
YoloV4 slower in INT8 than FP16 TensorRT	5	1491	June 5, 2021
Migrating INT8 calibration from TensorRT 6 to TensorRT 7 in YoloV3 and YoloV4 failed TensorRT tensorrt , jetson-inference	9	1523	December 28, 2020
Tensorrt Conversion TensorRT	2	88	November 30, 2024
Converting Yolo model to TensorRT format without ONNX conversion Jetson Projects tensorrt , yolo , pytorch , cudnn , inference-server-triton , deepstream	3	202	February 10, 2025
Darknet YoloV4-tiny model in TensorRT 8 inference TensorRT tensorrt , onnx	7	2241	October 22, 2021
Little performance difference between int8 and fp16 on RTX2080 TensorRT	4	2569	July 5, 2021
Yolov4 TensorRT slower than Yolov4 darknet TensorRT	6	3441	September 1, 2020

YoloV4 int8 conversion issue

Description

Environment

Related topics