onnx to tensorrt : Assertion failed: scales.is_weights() && "Resize scales must be an initializer!"

[yuzhj]

trafficstars

Here is the error log Here is my environment cuda 11.6 TensorRT 8.2.3.0 ONNX 1.12 ONNXRuntime 1.8.1 Do you have any advice? Thanks in advance

[CoinCheung]

What are the steps to reproduce this error?

[yuzhj]

step 1

python tools/export_onnx.py --config configs/bisenetv1_coco.py --weight-path pretrained/coco_stuff/model_final_v1_coco_new.pth --outpath ./model_v1.onnx

step 2

./tensorrt/build/segment compile model_v1.onnx model_v1.trt

I faced this error when I convert the model from onnx to trt @CoinCheung

[CoinCheung]

Maybe switch to tensorrt 8.2.5 and try again? Or use cuda 11.3?

[yuzhj]

Maybe switch to tensorrt 8.2.5 and try again? Or use cuda 11.3?

same error on tensorrt 8.2.5 and cuda 11.3

[CoinCheung]

That is weird, it works on my platform. Did you make any changes to the code? Would you please align the cudnn version as mine and try again?

[CoinCheung]

I am closing this, since this is not activate for a long time. Never mind, you can still open new issues, if you would like.