domain_classification

A binary domain classifier based on positive labels and an active learning algorithm.

For more information, go to the documentation pages.

Dockerfile explanation

The dockerfile contains all the instructions to generate a docker image, a template to build a docker container which contains the application.

• Base image: the base image for the application is ubuntu+python
• Directories architecture: the application will be in the /app directory inside the container, and the /shared_data directory will contain shared data with host.
• Application: copy application info to current directory (/app)
• Execution: the command in CMD is the default option when executing the container and can be overwritten by the user.

Create Docker image with WordList manager

Use the following command to create a docker image.

The flag -t specifies the name of the image that will be created with an optional tag (for example its version).

docker build <-t NAME:tag> <Dockerfile location>

Example:

docker build -t dom-class .

• The name of the image in this case is dom-class, with no specific version.
• The location of the Dockerfile is the current directory.

Run image container

To run a container with the previous configuration, the following command is needed:

docker run <--rm> -i <--name CONTAINER-NAME> -v path/to/local/data:/data/data IMAGE-NAME

Flags:

• --rm: remove the container when execution ends. (optional)
• -i: set interactive mode. It is required to use standard input
• --name: a name for the container. (optional)
• v: volume binding. It maps a local directory to a directory inside the container so that local files can be accessed from it. The format is: /absolute/path/to/local/dir:/absolute/path/to/container/dir

Example

Execute container and access

Create a container and access the CLI, as specified in Dockerfile

docker run --rm -i --name container-name -v path/to/local/data/:/shared_data/ dom-class

Show help menu

docker run --rm -i --name container-name -v path/to/local/data/:/shared_data/ dom-class --help

Execute task

docker run --rm -i --name container-name -v path/to/local/data/:/shared_data/ dom-class python main_dc_single_task.py --p /shared_data/projects/docker_proj --source /shared_data/datasets --task 

Alternatively, the command to execute the CLI can be executed first, and then the main command:

python run_dc_task.py --p /shared_data/projects/docker_proj --source /shared_data/datasets --task