README.md

object_detection1

How to Run

1. set target model (yolo11n)

   git clone https://github.com/ultralytics/ultralytics
   cd ultralytics
   pip install ultralytics
   

2. generate .onnx from timm model

   cd ..
   python onnx_export.py     # pytorch model -> onnx(fp16 or fp32)
   

3. build tensorrt model and run

   python onnx2trt.py
   

• fp16
  [TRT_E] 1000 iterations time: 3.4358 [sec]
  [TRT_E] Average FPS: 291.05 [fps]
  [TRT_E] Average inference time: 3.44 [msec]
  GPU mem : 164M

• ultralytics