Skip to content

Latest commit

 

History

History
50 lines (46 loc) · 971 Bytes

README.md

File metadata and controls

50 lines (46 loc) · 971 Bytes

Zaloai_landmark


What is it?


This project helps people submit the simplest way for the Zalo AI landmark identification challenge.

What does it contain?


  • Jupyter Notebook (Anaconda 3)
  • Keras
  • Tensorflow
  • Python 2 and 3
  • cuDNN (gpu version)

How to install it?


On Ubuntu

  1. Open a terminal
  2. Install docker
$ sudo apt-get install docker.io
  1. Clone file from repository
$ git clone https://github.com/dukn/zaloai_landmark.git
  1. Build the docker With CPU only
$ cd zaloai_landmark
$ sudo docker build -t zaloai/landmark -f Dockerfile.cpu . 

With GPU version

  • install nvidia docker in here
$ cd zaloai_landmark
$ sudo docker build -t zaloai/landmark -f Dockerfile.gpu . 

How to user it?


With CPU version

$ sudo docker -it -p 8888:8888 zaloai/landmark

With GPU version

$ sudo docker --runtime=nvidia -it -p 8888:8888 zaloai/landmark