Need TensorRT Engine for PeopleNet (A100 + TRT 10.13)

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

I need a pre-built TensorRT engine file for PeopleNet ResNet34 INT8. I have the .etlt model but can’t build the engine locally.

Request:

  • Convert resnet34_peoplenet_int8.etlt → .engine file
  • Target: A100 GPU, TensorRT 10.13, CUDA 12.2

Build command (if you have DeepStream/TAO):
tao-converter -k nvidia_tlt -d 3,544,960
-o output_cov/Sigmoid,output_bbox/BiasAdd
-e resnet34_peoplenet_int8.engine -t int8
-c resnet34_peoplenet_int8_calib.txt
models/resnet34_peoplenet_int8.etlt

Takes ~5 min with DeepStream. Can you help?

Thanks!

4

Hi @jialiang.wu, TensorRT engine files are hardware specific and optimized for the particular GPU architecture where they’re built.
It’s best practice to generate them on the target deployment system to ensure optimal compatibility and performance. you build command seem correct to me, and they should build the .engine file successfully. If something is failing, please share the logs so I can see where I might help you troubleshoot.

Thank You

5

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