1. Functional Features of the Service
1.1. The Service allows Customer to manage the machine learning lifecycle using cloud-based resources on the Platform. 1.2. The Customer is provided with the following options as part of the Service: 1.2.1. tracking and managing machine learning experiments; 1.2.2. recording and comparing experiment parameters and results; 1.2.3. storing and managing machine learning models and artifacts; 1.2.4. other functional features of the Service. The full list of functional features is available to Customer on Site and / or in Management Console. 1.3. Service Level is determined for the Service in accordance with SLA.2. Pricing
2.1. The use of the Service is chargeable. 2.2. Service Fee for the Service is determined as per Service Rates and billing units specified on Site. Nebius may change Service Rates as prescribed by Agreement. 2.3. Service Fee for the Service depends on: 2.3.1. computational resources allocated to the Managed MLflow Cluster based on the selected Resource Preset; 2.3.4. the Object Storage volume of the Artifacts Storage 2.4. Nebius may from time to time provide Customer with additional functional features of the Service for an additional fee.3. Definitions
Managed MLflow Cluster means a dedicated set of computing resources provisioned and managed by Nebius that hosts the MLflow services, including the MLflow Tracking Server, Metadata DB, and related components. Artifacts Storage means the designated object storage bucket where MLflow stores artifacts generated during the machine learning lifecycle, including models, datasets, plots, and other files. Experiment means a series of computational tasks, code executions, and/or data processing operations conducted by Customer within the Service, including input data, processing code, output results, and associated metadata. Resource Preset means a predefined configuration of computational resources (including CPU, memory, replicas, and other computing parameters) that can be allocated to a Managed MLflow Cluster. Artifacts Storage means the designated object storage bucket where MLflow stores artifacts generated during the machine learning lifecycle, including models, datasets, plots, and other files.Web address: https://docs.nebius.com/legal/specific-terms/managed-mlflow Publication date: March 4, 2025
Effective date: March 18, 2025 Previous version of the document: https://docs.nebius.com/legal/archive/specific-terms/managed-mlflow-20240830