Replies: 4 comments
-
Hello @MyraBaba we currently don't have a ready dockerfile but it is easy to create one. Feel free to open a PR and we will accept it. You can create a docker file that starts from a
pip install sahi
pip install pytorch=1.10.2 torchvision=0.11.3
pip install yolov5==6.1.3
pip install mmcv-full==1.5.3 -f https://download.openmmlab.com/mmcv/dist/cu113/torch1.10.0/index.html pip install mmdet==2.25.0
pip install detectron2 -f https://dl.fbaipublicfiles.com/detectron2/wheels/cu113/torch1.10/index.html
pip install transformers timm |
Beta Was this translation helpful? Give feedback.
-
@fcakyon to reduce to friction for new comer it will be very useful a fully working docker image.
now gives segmentation fault error(core dumped) |
Beta Was this translation helpful? Give feedback.
-
Hi @MyraBaba, can you please provide your platform information (OS, python, sahi versions, etc.) ? and also is that possible you can create a minimal reproducible example with some small video (so that we can try to recreate the problem you encounter) ? |
Beta Was this translation helpful? Give feedback.
-
Ubuntu 20.04 , 2080rtx ti , latest sahi , python3.8 will try findout. Additionally; when its working ı noticed that sahi is very slow , but gPU is utilzed %20.. FYI |
Beta Was this translation helpful? Give feedback.
-
Hi,
do you have docker image for easy test and fast cold start ?
Beta Was this translation helpful? Give feedback.
All reactions