yoloV8 quantisation fails

Description

I am trying to built an int8 model from a yoloV8.onnx following the instructions here:

This uses a 1000 calibration images to generate a calibration file and then is supposed to build the engine, but errors out with:

deepstream-app -c deepstream_app_config.txt

ERROR: [TRT]: 10: Could not find any implementation for node /0/model.22/Range.
ERROR: [TRT]: 10: [optimizer.cpp::computeCosts::3869] Error Code 10: Internal Error (Could not find any implementation for node /0/model.22/Range.)

The calibration file is created, but then the engine is not saved.
The config files for this are attached.

Environment

TensorRT Version: 8.6.2.3
GPU Type: Orin AGX 64 Gbyte
Nvidia Driver Version:
CUDA Version: 12.2.140
CUDNN Version: 8.9.4.25
Operating System + Version: Ubuntu 22.04 LTS
Python Version (if applicable): 3.10.12
TensorFlow Version (if applicable):
PyTorch Version (if applicable):
Baremetal or Container (if container which image + tag):

Relevant Files

deepstream_app_config.txt (907 Bytes)
config_infer_primary_yoloV8_int8.txt (680 Bytes)

Steps To Reproduce

Please include:

  • Exact steps/commands to build your repro
  • Exact steps/commands to run your repro
  • Full traceback of errors encountered

Hi @felix.darvas ,
Apologies for the delay,
Please raise teh concern to Deepstream Forum.

Thanks