Tensorrt3: How to set shape_node of uff.graph.conv_transpose?

nicklhy · February 9, 2018, 3:34am

Hi, I am trying to convert a mxnet model with deconv layer into UFF format. Currently, everything works fine if the network does not include any deconv layers (i.e. resnet, inception, alexnet).

But when converting the a CNN with deconv layers, tensorrt raise an error like below when building the engine.

python3: uff/UffParser.cpp:611: std::vector<long int> getOutputShapeSpatial(const Constant&, const Fields&, std::vector<int>): Assertion `inputOrder.size() == outShapeUnordered.size()' failed.

I guess the problem might be caused by the “shape_node” arg in “uff.graph.conv_transpose”. Currently, I create a constant node containing the output shape in numpy.ndarray format and use its name as the “shape_node” arg.

Can anyone provide a detailed example about creating a deconv layer in UFF format?

Topic		Replies	Views
[TensorRT] ERROR: UFFParser: Output shape of UFF ConvTranspose is wrong GPU-Accelerated Libraries	0	822	February 8, 2018
TensorRT3 gives a wrong shape for Transposed Convolution Jetson TX2	7	1903	October 18, 2021
TensorRT 3 Wrong output shape with ConvTranspose and Tensorflow TensorRT	4	996	August 31, 2018
Output shape of UFF ConvTranspose is wrong TensorRT	2	644	February 21, 2020
UFFParser error with conv2d_transpose TensorRT	0	826	June 19, 2019
Converted Tensorrt model has different output shape from Tensorflow model TensorRT	0	393	June 3, 2019
Parser error: Output shape of UFF ConvTranspose is wrong Jetson TX2	4	657	October 18, 2021
uff.model.exceptions.UffException: Const node conversion requested, but node is not Const TensorRT	8	1935	July 17, 2020
UFFParser reshape error TensorRT	1	1056	December 13, 2018
[TensorRT5.0] Parsing the converted UFF format InceptionV3 model failed, which is converted from TF frozen model TensorRT	4	1007	July 5, 2019

Tensorrt3: How to set shape_node of uff.graph.conv_transpose?

Related topics