Closed ttanzhiqiang closed 3 years ago
Hi, thank you for sharing. What does the last column mean? And your inference time seems too long for TensorRT. In our TRT model, YOLOX-S runs with avg 2.33 ms on a 2080ti.
this times include proprecess:8.70ms and forward:5.62ms and reprocessing:1.72ms the last column is GPU memory
look https://github.com/ttanzhiqiang/onnx_tensorrt_project