8000 GitHub - Sandro295/CutLER: Code release for "Cut and Learn for Unsupervised Object Detection and Instance Segmentation"
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Code release for "Cut and Learn for Unsupervised Object Detection and Instance Segmentation"

Notifications You must be signed in to change notification settings

Sandro295/CutLER

 
 

Repository files navigation

Homework

Here i am trying to build and/or run Docker container with someone else's Deep Learning repository.

Based on Facebook's CutLER

How to run Docker image

I have already built and pushed an Image to DockerHub

All you need to do is have Docker daemon running

On WSL

sudo dockerd

Then pull my image

If your CPU is x86

docker pull sandro295/cutler:multi-arch

If your CPU is ARM64

docker pull sandro295/cutler:arm64

And run the container

docker run sandro295/cutler:multi-arch

It will run a demo.py on your CPU twice then it will run a test which will compare two generated images with what i got on my machine. If they are similar it will print "Test passed".

After that the container will stop and output files will be availiable for copying

docker ps -a
docker cp <container_id>:/home/docker_user/images/outputs ~/<your_folder>

How to build Docker image

If you are brave enough you can try to build an image all by yourself.

Firstly clone this repository recursively

git clone https://github.com/Sandro295/CutLER.git --recursive
cd CutLER

Make sure you have Docker installed and 50DC your Docker daemon is running
Then build an image

docker build -t cutler .

You can try to run it with

docker run cutler

About

Code release for "Cut and Learn for Unsupervised Object Detection and Instance Segmentation"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 98.5%
  • Shell 1.3%
  • Dockerfile 0.2%
0