EfficientNet B0 from onnx to trt

hi ,
I am trying to convert the efficientnet B0 to trt from this tools https://github.com/NVIDIA/TensorRT/tree/main/samples/python/efficientnet
python3 build_engine.py --onnx efficientnet-b0.onnx --engine efficientnet_INT8.engine --precision int8 --calib_input ~/calibration/ --calib_cache ~/calibration.cache --calib_preprocessor V2

[TRT] [E] 4: [standardEngineBuilder.cpp::initCalibrationParams::1417] Error Code 4: Internal Error (Calibration failure occurred with no scaling factors detected. This could be due to no int8 calibrator or insufficient custom scales for network layers. Please see int8 sample to setup calibration correctly.)

may I know is there any missing parameter in this cmd ?

my usage model is what is the method to deploy a efficientnet B0 int 8 in deepstream ?
the onnx converted via this method ,

thanks

Hi,

Could you help to check if the fp16 mode works?

Thanks.

Hi,

Based on the output above, there is no error bug just some warning.
Do you generate the engine correctly?

Thanks.

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