تخطَّ إلى المحتوى

mlflow

هذا المحتوى غير متوفر بلغتك بعد.

ML experiment tracking, model registry, and artifact storage. Free: MIT licensed.

Terminal window
nself plugin install mlflow

Runs an MLflow tracking server backed by your Postgres database. Track experiments, compare model runs, log parameters and metrics, and manage the model registry. Integrates with Python ML workflows via the MLflow client.

Env VarDefaultDescription
MLFLOW_PORT5000MLflow server port
MLFLOW_ARTIFACT_ROOT/mlflow/artifactsArtifact storage path
PortPurpose
5000MLflow UI and REST API

0 plugin-managed tables: MLflow manages its own schema in Postgres directly via SQLAlchemy.

RouteTarget
mlflow.{base_domain}MLflow UI

Requires Postgres (always running in ɳSelf). Optionally uses MinIO for artifact storage, set MLFLOW_ARTIFACT_ROOT=s3://... and configure MinIO credentials.