Kshitij Srivastava

icon indicating a website link https://www.linkedin.com/in/kshitijsri

icon company Aurora icon location Pittsburgh icon location SW Engg - Aurora. Ex-Uber-ATG, Ex-ORNL. Working on optimizing DL models for inference on different H/W

Results 5 comments of Kshitij Srivastava

Error when running the example on the repository's homepage.

You are running your model and the input tensor on a cpu. it should be on a gpu

Why is the trt model file type .pth?

TRT engine is a black box. Netron wont work with a trt engine

deal the outputs cost more time, why?

You are explicitly copying tensors from gpu to cpu, hence , the cost of transfer

deal the outputs cost more time, why?

I mean if you really want to copy tensor from gpu to cpu, then you will have to incur some transfer time

torch2trt will impact the normal pytorch?

Tried this example with PR #691 and there is no error.