TensorRT 7 INT8 quantization

AI & Data Science Deep Learning (Training & Inference) TensorRT
1

FOr TRT 7 INT 8, I created a calibrator, and loaded 1000 pics. After I make the files and then run ./yolox, it has this error, is anyone met this before???

image
image2830×78 39.7 KB

I printed the log and found this line code report error.
CUDA_CHECK(cudaMemcpyAsync(output, buffers[1], output_size * sizeof(float), cudaMemcpyDeviceToHost, stream));

2

Hi, Please refer to the below links to perform inference in INT8

Thanks!

4

Cheers, I found the problem. It is that I did not allocate the correct memory for the output.

5

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