TensorRT parse onnx failed with AveragePool layer

spolisetty · September 16, 2021, 5:55am

After our team investigating the model, found that avgpool is an invalid one (attaching a screenshot of the model)

The input to the averagepool is a 3D tensor of shape 1x196x768. According to the ONNX spec, the corresponding pooling window, stride, and pads should be 1D, but provided as 2D values here.

Using ONNX-GS we edited this node to follow the proper ONNX spec of averagepool. The network can now be run fully in TensorRT. Please find the model here.
https://drive.google.com/file/d/1mC46F84UYLv_-JJfA2bTEhYFQ89MQjFs/view?usp=sharing

Thank you.

Topic		Replies	Views
Error at AveragePool while converting ONNX --> TensorRT TensorRT tensorrt , ubuntu	1	593	November 17, 2022
Parse onnx file failed: Parameter check failed. condition:condition: allDimsGtEq(windowSize, 1) && volume(windowSize) < MAX_KERNEL_DIMS_PRODUCT(nbSpat TensorRT tensorrt	3	1008	May 17, 2022
TensorRT cannot parse ONNX model TensorRT	5	1914	June 18, 2020
Pytorch model convert to TensorRT engine failed TensorRT tensorrt , pytorch , onnx	5	1309	December 28, 2020
ONNX and tensorRT: ERROR: Network must have at least one output TensorRT	30	17295	October 6, 2020
Onnx to torchrt convertion error TensorRT tensorrt , cuda , onnx , tf-trt , jetson	3	1181	February 8, 2022
[TensorRT] ERROR: Network must have at least one output [TensorRT] ERROR: Network validation failed TensorRT tensorrt , cuda , onnx	10	2917	October 16, 2020
tensorRT inference unstable compared onnxruntime TensorRT	4	1453	May 4, 2021
Run out of memory when creating TensorRT engine from onnx model Jetson Xavier NX tensorrt	7	2930	October 18, 2021
Tensorrt from onnx ERR: atleast 4 dimensions are required for input TensorRT	11	4273	February 17, 2020

TensorRT parse onnx failed with AveragePool layer

Related topics