search

isarsoft / yolov4-triton-tensorrt

This repository deploys YOLOv4 as an optimized TensorRT engine to Triton Inference Server
http://www.isarsoft.com
Other
276 stars 63 forks source link

How to use FP16 #34

Closed 17702513221 closed 3 years ago

17702513221 commented 3 years ago

I watch #define USE_FP16 ,but the model plan is TYPE_FP32

philipp-schmidt commented 3 years ago

Hi, this is only the input data type. The actual computation will be FP16 which you can see in performance increase. You can change the input data type if you need to, it is a parameter of the createEngine function.