How to Evaluate YOLOv7 Model Performance on NVIDIA Jetson?

• Hardware Platform (Jetson / GPU) Jetson Orin NX 16GB
• DeepStream Version Deepstream-6.2
• JetPack Version (valid for Jetson only) Jetpack 5.1
• TensorRT Version 5.1
• OpenCV without CUDA

I’m working on a YOLOv7 model and got its accuracy rating (mAP value) on my server. After converting the model to ONNX format, I upload it to an Jetson and use DeepStream to create an engine file. I can choose between different precision levels (FP32, FP16, INT8), which significantly affects speed. How can I measure performance using the mAP value after converting the model into the engine file?

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

You can refer to the test_coco_map.py to get the mAP.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.