MLflow Model Writer Module

This module uploads trained models to the MLflow server.

Parameter

Type

Description

Example Value

Default Value

conda_env string Conda environment for the model “path/to/conda_env.yml” [Required]
databricks_permissions dictionary Permissions for the model See Below None
experiment_name_formatter string Formatter for the experiment name “experiment_name_{timestamp}” [Required]
model_name_formatter string Formatter for the model name “model_name_{timestamp}” [Required]
timestamp_column_name string Name of the timestamp column “timestamp” timestamp

Key

Type

Description

Example Value

Default Value

read array List of users with read permissions [“read_user1”, “read_user2”] -
write array List of users with write permissions [“write_user1”, “write_user2”] -
Copy
Copied!
            

{ "model_name_formatter": "model_name_{timestamp}", "experiment_name_formatter": "experiment_name_{timestamp}", "conda_env": "path/to/conda_env.yml", "timestamp_column_name": "timestamp", "databricks_permissions": { "read": ["read_user1", "read_user2"], "write": ["write_user1", "write_user2"] } }

Previous FromControlMessage Module
Next Batch Data Payload Module
© Copyright 2023, NVIDIA. Last updated on Feb 2, 2024.