ERRORINVALID_ARGUMENT: Cannot find binding of given name

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

Description

I am converting an .onnx model to a TensorRT format on google colab using the command:

!./makeCudaEngine -i ../../ScaledYOLOv4/yolov4-p5.onnx -o yolov4-p5.trt

the command does save a yolov4-p5.trt but with the following output

ERRORINVALID_ARGUMENT: Cannot find binding of given name: conf
ERRORINVALID_ARGUMENT: Cannot find binding of given name: cls
ERRORINVALID_ARGUMENT: Cannot find binding of given name: bbox
save yolov4-p5.trt

Environment

Google Colab

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2020 NVIDIA Corporation
Built on Mon_Oct_12_20:09:46_PDT_2020
Cuda compilation tools, release 11.1, V11.1.105
Build cuda_11.1.TC455_06.29190527_0
Ubuntu 18.04.5 LTS \n \l

How can I fix this? Thanks!

4

Hi @jcdulo ,
Can you please share your onnx model and reproducible script, so that we can try the same at our end.

Thanks!

5

Thank you for replying @AakankshaS.

I am following the steps in this repository:
https://github.com/talebolano/TensorRT-Scaled-YOLOv4

The error comes up at step 3. Convert onnx model to try model

./makeCudaEngine -i ../../ScaledYOLOv4/yolov4-csp.onnx -o yolov4-csp.trt

and affects even step 4.Test

I am including a link to my google colab as well for your review.
Google Colab Notebook

Thank you.

6

Hi @jcdulo,

Sorry for the delayed response. We are unable to run google colab notebook successfully.
We request you to please share ONNX model to try from our end.

Meanwhile you can also try trtexec to convert ONNX model to trt.
In case you are still facing issue, request you to share the trtexec “”–verbose"" log for further debugging

For your reference,
https://github.com/NVIDIA/TensorRT/tree/master/samples/opensource/trtexec

Thank you.