Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

72 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Prerequisites

  1. Install Docker

  2. Download compose.yaml

  3. Make sure that the contract and the sql file with the query to the Aktin DWH are on the same host, you can find examples here

  4. Create .env in the same folder as compose.yaml and store your setup information, e.g.

    • Docker compose automatically reads the contents of .env
    # Aktin DWH connection info
    #DB_HOST=my-host # do not set if db is hosted on localhost
    DB_PORT=5432
    DB_USER=i2b2crcdata
    DB_SCHEMA=i2b2
    # orchestrator information
    # optional, only needed if file path differs from ./contract.yaml
    MELD_CONTRACT_FILE=$HOME/MELD/examples/nn/resources/contract.yaml
    # optional, only needed if file path differs from ./query.yaml
    MELD_QUERY_FILE=$HOME/MELD/examples/nn/resources/query.sql
  5. In the same folder also create a file secrets.txt and fill it with the db password

  6. Pull MELD image docker compose pull

Load inference runtime image

Before the first inference run, the inference runtime image must be pulled.

  1. Open a terminal and navigate to the folder containing the compose.yaml
  2. Run docker compose run meld pull

Run inference job

  1. Open a terminal and navigate to the folder containing the compose.yaml
  2. Run docker compose run meld run

Delete runtime image

1.Open a terminal and navigate to the folder containing the compose.yaml 2 Run docker compose run meld delete

Get results

After a completed (successful or unsuccessful), you can find the output archive in ./jobs/<job_id>/outout/summarized_execution.zip. It contains

  • input/: all input files like input.csv and contract.yaml
  • logs/: logs of the job
  • output/: result files from the runtime container, like output.csv

About

Machine Learning Engine

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages