Demo the work of MLRun with Github based projects and automated CI/CD
in a client or notebook properly configured with MLRun and KubeFlow use the following lines:
from mlrun import load_project
# load the project from GitHub
url = 'git://github.com/mlrun/webhook-demo.git'
proj = load_project('/tmp/myproj', url)
print(proj.to_yaml())
# run the project main pipeline (build, data prep, train, deploy model)
pipeline = proj.run(arguments={}, artifacts_path='v3io:///users/admin/mlrun/kfp/{{workflow.uid}}/')
- Project spec (functions, workflows, etc)
- Local function spec (XGboost)
- Function code
- Workflow code (init + dsl)