Modellstadier

Introduktion till MLflow

Weston Bassler

Senior MLOps Engineer

Mjukvarumiljöer

Diagram över utvecklings-, staging- och produktionsmiljöer

MLflow-modellstadier

Tilldelas modellversioner
Fördefinierade stadier:
- None
- Staging
- Production
- Archived
Ett stadium i taget

Fördefinierade stadier

None

Stadiet None

Staging

Stadiet Staging

Production

Stadiet Production

Archived

Stadiet Archived

Övergång mellan stadier

transition-ui

# Import MLFlow Client
from mlflow import MlflowClient
client = MlflowClient()


# Transition to Staging
client.transition_model_version_stage(
        name="Unicorn",
        version=3,
        stage="Staging"
    )

Övergång till Staging för modellversion

# Transition to Staging
client.transition_model_version_stage(name="Unicorn", version=3, stage="Staging")

<ModelVersion: creation_timestamp=1679693470034, current_stage='Staging', 
description=None, last_updated_timestamp=1679699050734, name='Unicorn', 
run_id='a1454f2865e449f8835f38f71e53e547', run_link=None, 
source='./mlruns/1/a1454f2865e449f8835f38f71e53e547/artifacts/model', 
status='READY', status_message=None, tags={}, user_id=None, version=3>

Registrets användargränssnitt

staging-ui

Övergång till Production

# Transition to Production
client.transition_model_version_stage(name="Unicorn", version=3, 
                                      stage="Production")

<ModelVersion: creation_timestamp=1679693470034, current_stage='Production', 
description=None, last_updated_timestamp=1679699633297, name='Unicorn', 
run_id='a1454f2865e449f8835f38f71e53e547', run_link=None, 
source='./mlruns/1/a1454f2865e449f8835f38f71e53e547/artifacts/model', 
status='READY', status_message=None, tags={}, user_id=None, version=3>

Nu kör vi en övning!

Introduktion till MLflow