Open Daniel-Fundarek opened 7 months ago
Is it possible to export model using TensorRT to int8. If so how?
Yes, rtdetr can using tensorrt with int8.
For paddle model please reference this.
or
You should check nv trt doc or thirdpart tutorial for x -> onnx -> tensorrt. Perhaps this can help you .
x -> onnx -> tensorrt
Is it possible to export model using TensorRT to int8. If so how?