Closed XinnWang closed 4 years ago
The NVIDIA website has an example on how to convert PackNet to TensorRT, you can have a look here:
https://docs.nvidia.com/deeplearning/tensorrt/sample-support-guide/index.html#onnx_packnet
Thanks!
Hi, i tried the example, it works well. I am trying to increase the inference speed by INT8 mode, but the results of INT8 mode is wrong. Have you ever tried INT8 inference ? any suggestions? Thanks.
We still haven't tried INT8 inference, so unfortunately I cannot help you with that. I will leave this issue open for now, maybe someone else can provide more information.
Hi, i noticed that in your CVPR paper, you said:
So i try to use TensorRT to improve the performance:
However, step 2 failed, the log info is below:
i tried the DepthResNet and PackNet01 model, both failed at this ERROR:
Have you met this error before? Any suggestions? Thanks.