Description
I used TensorRT v8502 to convert a U-Net model on Orin, utilizing a pre-calibrated calib file with the MinMax method. However, when using trt.IInt8MinMaxCalibrator in the script to convert the model on Orin, an error occurred.
[11/19/2024-13:16:49] [TRT] [W] Network built for DLA requires kENTROPY_CALIBRATION_2 calibrator.
[11/19/2024-13:16:49] [TRT] [E] 4: [network.cpp::validate::2789] Error Code 4: Internal Error (DLA validation failed)
I tried manually changing the header in the calib file from TRT-8502-MinMaxCalibration to TRT-8502-EntropyCalibration2, which made the conversion succeed, but the inference results of the model were incorrect.
Environment
TensorRT Version: 8.5
GPU Type: orin
Nvidia Driver Version:
CUDA Version: cuda-11.4
CUDNN Version:
Operating System + Version: ubuntu 20.04
Python Version (if applicable): 3.8.10
TensorFlow Version (if applicable):
PyTorch Version (if applicable):
Baremetal or Container (if container which image + tag):
Relevant Files
Please attach or include links to any models, data, files, or scripts necessary to reproduce your issue. (Github repo, Google Drive, Dropbox, etc.)
Steps To Reproduce
Please include:
- Exact steps/commands to build your repro
- Exact steps/commands to run your repro
- Full traceback of errors encountered