morpheus.modules.mlflow_model_writer
Functions
mlflow_model_writer (builder) |
This module uploads trained models to the mlflow server. |
- mlflow_model_writer(builder)[source]
This module uploads trained models to the mlflow server.
- Parameters
- builder
mrc Builder object.
Notes
- Configurable Parameters:
conda_env (str): Conda environment for the model; Example:
path/to/conda_env.yml
; Default:[Required]
databricks_permissions (dict): Permissions for the model; See Below; Default: None
- experiment_name_formatter (str): Formatter for the experiment name;
Example:
experiment_name_{timestamp}
; Default:[Required]
- model_name_formatter (str): Formatter for the model name; Example:
model_name_{timestamp}
;
Default:
[Required]
- model_name_formatter (str): Formatter for the model name; Example:
timestamp_column_name (str): Name of the timestamp column; Example:
timestamp
; Default: timestamptimeout (float): Timeout for get requests.
- databricks_permissions:
read (array): List of users with read permissions; Example:
["read_user1", "read_user2"]
; Default: -write (array): List of users with write permissions; Example:
["write_user1", "write_user2"]
; Default: -