search

linghu8812 / tensorrt_inference

708 stars 208 forks source link

Segmentation fault #170

Closed Egorundel closed 1 year ago

Egorundel commented 1 year ago

Hello, I have the following problem:

Screenshot from 2023-07-06 12-12-10

Can you help me to solve this problem, please?

TensorRT Version: 7.2.2.3.

NVIDIA GPU: RTX 2070

CUDA Version: 11.1

Operationg System: Ubuntu 20.04

Egorundel commented 1 year ago

The problem was solved by replacing void *buffers[2]; in model.h on void *buffers[5];

and in model.cpp changed the line assert(engine->getNbBindings() == 2); to assert(engine->getNbBindings() == 5);

Do this if you use the NMS Plugin in the model :)

c7-biu commented 4 months ago

Hi! I have met this problem too! Could you please share the solutions with me? Thank you!

Egorundel commented 4 months ago

@c7-biu Attach your ONNX-model, I will try to understand your problem

c7-biu commented 4 months ago

I didn't use ONNX,I try to ues the tensorrtx to accelerate the yolov8,but the problem i met is similar with you. prolem

Egorundel commented 4 months ago

@c7-biu What do you have at the input and output of the model?

c7-biu commented 4 months ago

Is this it? ![Uploading 1.png…]()

I'm a novice at YOLO, so maybe I don't understand some of the technical words very well.