the results of 'giexec' tensorrt

Autonomous Machines Jetson & Embedded Systems Jetson TX2
1

Hi.
I’m trying to use tensorrt on tx2.
The command, ‘giexec’, at /usr/src/tensorrt/bin prints times as a results.
What does these times (numbers) mean?

I think these numbers are time for classification or forward pass. Right?

AND I have another question.
When I made batch size up (like 4 of other bigger number), the time became larger than when the batch size is 1(default size).
Why it happens? What dose it mean?

AND how to classify images using alexnet and tensorrt. I want to classify the image and check the classification result and consuming time.

2

Hi,

This issue duplicates to topic-1023330. Please check it for the answer:
[url]https://devtalk.nvidia.com/default/topic/1023330/jetson-tx2/batch-size-and-the-time-consuming/post/5206490/#5206490[/url]

Thanks.

1 Like
3

So …What does these times (numbers) mean?
The time run forward once?To recognize one picture?

4

Hi,

It is the result from profiler which profiles TensorRT execution layer by layer.
[url]https://docs.nvidia.com/deeplearning/sdk/tensorrt-developer-guide/index.html#profiling[/url]

Thanks.