Training with MLflow

Train and register a model while preserving data, parameters, metrics, and artifacts.

ingest data→validate→train→evaluate→register modellog MLflow run

Training with MLflow

Activity responsibilities

Activity	Responsibility
ingest-data	Execute the ingest data stage and publish its declared outputs for downstream activities.
validate	Execute the validate stage and publish its declared outputs for downstream activities.
train	Execute the train stage and publish its declared outputs for downstream activities.
evaluate	Execute the evaluate stage and publish its declared outputs for downstream activities.
register-model	Execute the register model stage and publish its declared outputs for downstream activities.
log-mlflow-run	Execute the log MLflow run stage and publish its declared outputs for downstream activities.

Inputs

• Versioned dataset
• Training configuration
• Runtime image

Outputs

• Model artifact
• Evaluation report
• MLflow run

Provenance to preserve

• Dataset and code digests
• Parameters and metrics
• Resource placement

Source bundle

The executable workflow and its implementation files are maintained in the separate AkôFlow examples repository. The documentation intentionally keeps code, container recipes, generators, and validation scripts out of the narrative page.

Open the complete source bundle ↗

Verified local execution

These captures and the output manifest were produced by the fixture's local Docker run and validator. They are published with the same bundle as the runnable files.

Workflow evidence

Execution evidence

Output evidence

Execution considerations

Use GPU capacity for training and CPU capacity for validation or registration. MLflow remains an external service; AkôFlow records the activity that communicates with it.

AkôFlow boundary: the engine schedules, deploys, executes, transfers data, and records evidence. The ML or agent framework remains an implementation choice inside each activity.