TensorRT Python INT8 calibration failure

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

I am trying to calibrate a caffe model to infer with INT8 precision.
I am following this guide: https://devblogs.nvidia.com/parallelforall/int8-inference-autonomous-vehicles-tensorrt/

The only change I made is changing the model to classification (googlenet, using argmax as post-processor), and using jpg for calibration dataset.

I get the following error:

[TensorRT] INFO: Calculating Maxima
[TensorRT] ERROR: Calibration failure occured with no scaling factors detected. Please see int8 sample to setup calibration correctly.
[TensorRT] ERROR: Failed to create engine

Can you provide any idea on why this is happenning?

Thanks in advance.

2

Seems that it is not supported yet.

3

Did you manage to solve your issue here.

4

I have the same problem
Are there any other INT8 pipeline sample for python？ I only find the C++ pipeline in the official samples.