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

DCN v2 compilation for RTX 3080 #204

Open
ghost opened this issue Apr 8, 2021 · 5 comments
Open

DCN v2 compilation for RTX 3080 #204

ghost opened this issue Apr 8, 2021 · 5 comments

Comments

@ghost
Copy link

ghost commented Apr 8, 2021

Hello,

I am not able to compile DCN v2. My configuraiton is the following:

conda 4.9.2
python 3.8.8
pytorch 1.8.1
Build cuda_11.2.r11.2/compiler.29618528_0
driver version: 460.39
CUDA Version 11.2

I have tried many DCN v2 forks to compile but I was not able succeed.
Any helpd would be appreciated :)

Thank you
Fatih.

@nbraem
Copy link

nbraem commented Jul 16, 2021

You can switch out to a pure pytorch DCN, e.g.:
https://github.com/developer0hye/Simple-CenterNet/blob/main/models/dcn.py#L5

@nidetaoge
Copy link

You can try downgrading pytorch to 1.7.

@Xuguozi
Copy link

Xuguozi commented Aug 30, 2021

my pytorch is 1.7,and it works for compling

@youonlytrackonce
Copy link

@bryanbocao
Copy link

Try using Docker:
https://github.com/bryanbocao/CenterTrack/blob/master/readme/Docker.md
#290

Run CenterTrack using Docker

DCNv2 Compilation Issue Fixed

docker pull bryanbocao/centertrack
docker run -d --ipc=host --shm-size=16384m -it -v /:/share --gpus all --network=bridge bryanbocao/centertrack /bin/bash
docker ps -a

Check <CONTAINER_ID>:

CONTAINER ID   IMAGE                    COMMAND                  CREATED          STATUS                       PORTS                NAMES
89bb79551ccb   bryanbocao/centertrack   "/usr/local/bin/nvid…"   49 seconds ago   Up 38 seconds                6006/tcp, 8888/tcp   competent_northcutt
docker exec -it <CONTAINE_ID> /bin/bash
docker exec -it 89bb79551ccb /bin/bash

Inside the container:

cd /root/CenterTrack/src/lib/model/networks/DCNv2
python3 setup.py build develop
cd /root/CenterTrack/src/
python3 demo.py tracking,ddd --load_model ../models/nuScenes_3Dtracking.pth --dataset nuscenes --pre_hm --track_thresh 0.1 --demo ../videos/nuscenes_mini.mp4 --test_focal_length 633

The DCNv2 is built from MatthewHowe_DCNv2. Thanks MatthewHowe for the contribution.
Then you should be able to see the results.

Go to /share to transfer data between the container and your root system.

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

5 participants