Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

使用TensorRT推理STARK-Lightning onnx模型 #87

Open
Stoooner opened this issue May 9, 2022 · 5 comments
Open

使用TensorRT推理STARK-Lightning onnx模型 #87

Stoooner opened this issue May 9, 2022 · 5 comments

Comments

@Stoooner
Copy link

Stoooner commented May 9, 2022

1. 测试平台

  1. ubuntu18.04(x86_64)
  • 显卡: RTX 2060;
  • cuda_version: 10.2.89;
  • cudnn_version: 8.2.2.26;
  • opencv_version: 4.5.3;
  • gcc_version: 7.5.0;
  • onnxruntime_version: onnxruntime-linux-x64-gpu-1.6.0;
    -TensorRT_version: 8.0.1.6

2. 实际效果

通过对STARK-Lightning的onnx文件进行修改并重新导出之后,通过编写TensorRT代码成功编译并进行了执行,通过与ONNXRunTime(C++)的推理效果进行了简单的比较,效果如下:

  • TensorRT效果(FPS大体在260左右):
    微信图片_20220509084111
  • ONNXRunTime(C++)效果(FPS大体在220左右):
    微信图片_20220509084121

3. Jetson TX2(大疆Dji 妙算2)

刚刚在TX2上测试了一下,能够稳定在60 FPS左右,具体细节后续补充:
IMG_20220510_095235.jpgIMG_20220510_095240.jpg
IMG_20220510_094903.jpg

@jcyhcs
Copy link

jcyhcs commented May 19, 2022

@Stoooner
hi,professor:
i also try to convert onnx model to TensorRT,use trtexec, but i got an error like this:
ERROR: builtin_op_importers.cpp:980 In function importGather:
[8] Assertion failed: data.getType() != nvinfer1::DataType::kBOOL

oh,can you tell me how to modify this problem?
please help!

@Stoooner
Copy link
Author

@Stoooner
hi,professor:
i also try to convert onnx model to TensorRT,use trtexec, but i got an error like this:
ERROR: builtin_op_importers.cpp:980 In function importGather:
[8] Assertion failed: data.getType() != nvinfer1::DataType::kBOOL

oh,can you tell me how to modify this problem?
please help!

hi,既然你看得懂中文,那我就中文发了哈。我的邮箱是2582814014@qq.com,你可以通过邮箱发一下你的微信,我加你,我们微信聊,不然用github的方式效率太低了

@YukikkoNano
Copy link

请问这个是不通过deepstream的吗?我想在jetson上运行自己魔改的yolo一直没成功过……

@Stoooner
Copy link
Author

请问这个是不通过deepstream的吗?我想在jetson上运行自己魔改的yolo一直没成功过……

不通过deepstream,yolo的我也运行起来了,邮箱2582814014@qq.com,发邮件加微信详聊呗。。。

@zzxaijs
Copy link

zzxaijs commented Jul 31, 2022

导出的onnx,修改了什么地方,可否分享一下?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants