This repository is built to conduct an experimental evaluation of TensorRT's optimization capabilities on GPUs of various ranges, from embedded systems to desktop computers, providing quantitative data on its advantages and limitations. The findings offer guidelines for optimizing customized algorithms in scientific and industrial applications beyond standardized benchmarking examples.
See the Wiki for full documentation, examples, operational details and other information.
To build in a Jetson device, refer to the Wiki
- Linux ( based on Ubuntu 22.04 LTS )
- virtualenv
- CUDA 12.4
- ultralytics ( only to test on yolov8 )
$ git clone [email protected]:Juanx65/ArtTRT.git
$ cd ArtTRT/
$ virtualenv env
$ source env/bin/activate
$ pip install --no-cache-dir -r requirements.txt
Refer to workflow.ipynb.