NVIDIA Developer Forums

about custom layer error

Accelerated Computing GPU-Accelerated Libraries

yzyoyoyo February 8, 2018, 10:06am 1

I use python api to create a custom layer,

and the “import tensorrtplugins” is successful.

However, it is wrong in assert(engine.get_nb_bindings() == 2),
what’s it mean?

Topic		Replies	Views	Activity
Assertion `engine->getNbBindings() == 2' failed when after doing INT8 calibration Jetson Xavier NX tensorrt	4	2112	September 27, 2021
TensorRT custom layer help TensorRT	4	501	June 11, 2020
TensorRT add_plugin() in Python GPU-Accelerated Libraries	1	1058	January 29, 2018
Documentation for TensorRT 3.0 GPU-Accelerated Libraries	0	522	December 6, 2017
TensorRT3 gives wrong result for alphago model GPU-Accelerated Libraries	0	456	January 16, 2018
how to create custom layer in TensorRt by python GPU-Accelerated Libraries	0	584	February 5, 2018
Implement Plugin Layer with support of FP16 mode TensorRT	0	1018	April 26, 2019
TensorRT 2 EA and concat layers GPU-Accelerated Libraries	0	822	June 13, 2017
Is Plugin API supported by onnx parser? TensorRT tensorrt	4	452	October 12, 2021
int8 calibration,meet error get_batch() takes 2 positional arguments but 3 were given TensorRT	5	2235	April 8, 2020