Animegan_Paddle

Docker version API for AnimeGANv2!

The webapp is deployed Divio-Online here - https://animegan-paddle.us.aldryn.io/

The webapp is deployed Divio-Test - https://animegan-paddle-stage.us.aldryn.io/

The webapp is deployed Heroku here - https://apaddle-demo.herokuapp.com/

---

What is this?

• Original address of the projectAnimeGANv2
• The project used by this versionAnimegan2-Pytorch

Explain

AnimeGANv2(Look at the picture)

     

Compiled project on hub.docker.com

• Normal version
• GPU version
• Heroku version

Build

Make sure you have docker installed

1. Clone the Animegan repository:

   git clone https://github.com/LiteraturePro/Animegan_Paddle.git
   cd Animegan_Paddle
   

2. Input command to build image：

   docker build -t animegan .
   

   • I also provided the compilation command for Heroku, just replace the last command of dockerfile file with each other,
   • For general

   CMD exec gunicorn --bind 0.0.0.0:8080 --workers 1 --threads 8 --timeout 0 app:app
   

   • For heroku

   CMD exec gunicorn --bind 0.0.0.0:$PORT --workers 1 --threads 8 --timeout 0 app:app
   

3. Running image (You can specify the running port yourself)：

   docker run -p 8080:8080 animegan
   

Install

Make sure you have docker installed

I have built the image and can install it directly. The installation command is as follows(You can specify the running port yourself)：

• For general

  docker pull literature/animegan-paddle:latest
  docker run -p 8080:8080 literature/animegan-paddle:latest
  

• For heroku

  docker pull literature/animegan-paddle:heroku
  

  Please see the specific tutorial for installing container application in heroku

• For Aliyun Serverless

  docker pull literature/animegan-paddle:sf
  

  Please see the specific tutorial for installing container application in Aliyun Serverless Now your service has started to run, but it runs on the local port. If you need to realize the external network call, you need to act as an agent to proxy the service to your domain name，

Use

The call I have shown is based on the agent I have done. If you need to call it, you need to do it yourself

• provided that you have installed docker. After you deploy correctly, both GET and POST requests can be accessed. The actual display is as follows
  • Interface: http://your domain/api or http://127.0.0.1:8080/api can be accessed.
  • Parameter: image value: a picture
  • Return value: the base64 data stream after processing the image

Other

Thanks for the work of the original author and the revised author. If you like, please give a star.