ERROR: ModelImporter.cpp:288 In function importModel issue

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

Input filename: models/onnx/onnx_model_gcanet.onnx
ONNX IR version: 0.0.4
Opset version: 11
Producer name: pytorch
Producer version: 1.3
Domain:
Model version: 0
Doc string:

WARNING: ONNX model has a newer ir_version (0.0.4) than this parser was built against (0.0.3).
While parsing node number 0 [Conv]:
ERROR: ModelImporter.cpp:288 In function importModel:
[5] Assertion failed: tensors.count(input_name)
[E] Failure while parsing ONNX file

Because I use “F.interpolate(edge2_fea, size=(h, w), mode=‘bilinear’, align_corners=False)”, pytorch must be 1.3version otherwise some issue happens:
UserWarning: ONNX export failed on ATen operator upsample_bilinear2d because torch.onnx.symbolic_opset9.upsample_bilinear2d does not exist

But tensorRT issue can fix by downgrading pytorch to 1.2.

Can you tell me how to modify them? Thanks.

same issue:
https://devtalk.nvidia.com/default/topic/1066078/-5-assertion-failed-tensors-count-input_name-/?offset=1

2

Hi,

You can use TRT 7, it supports PyTorch 1.3.
Please refer below link for more details:
https://docs.nvidia.com/deeplearning/sdk/tensorrt-release-notes/tensorrt-7.html#tensorrt-7

Thanks