NVIDIA-AI-IOT / yolo_deepstream

yolo model qat and deploy with deepstream&tensorrt
Apache License 2.0
550 stars 139 forks source link

How to dynamically modify the model at runtime? For example, when you use yolov5 for one time, you will automatically switch to another model yolov7 after a period of time? #28

Closed flyingfish0918 closed 1 year ago

wanghr323 commented 1 year ago

Errr..... You can init two models with tensorrt, first using yolov5, after 1 time inference. use yolov7 to do inference instead. control with "if-else" in your program. That sounds strange, But you can do it if you like.......