Skip to content

Latest commit

 

History

History
49 lines (30 loc) · 1 KB

README.md

File metadata and controls

49 lines (30 loc) · 1 KB

README

Docker container for Pysangamam Workshop : Computer Vision through the Ages.

You can install Docker CE for Linux or MacOS or use dockertools on Windows.

Importing the Image

Import the image as:

cat /path/to/cvi-pysng.tar | docker import  - cvi-pysng

Run the image as:

docker run -it --name cvi -p 8888:8888 cvi-pysng bash

Mounting a volume is optional, but you may find it easier to make changes in files (vim is bundled in the container too).

Now, in your container,

  • You can access the bottleneck scripts at keras-bottlenecks
  • You can run a jupyter notebook for the code snippets:
jupyter notebook --ip=0.0.0.0 --allow-root

This should redirect to port 8888 (on localhost). If the port is blocked, try another port with the --port flag.

Building the Image

Build the image as:

docker build -t cvi-pysng -t p .

Run the image as:

docker run --it --name cvi -p 8888:8888 -v /some/mount/:/pysangamam cvi-pysng bash

The remaining steps are the same.