You are using an Azure Databricks cluster for training your ML model.
After successfully training your model built using the scikit-learn framework, you want to register it to the backend tracking server, using the following code:
# register model mlflow.sklearn.log_model(model, artifact_path = "trained_model", registered_model_name = 'my_trained_model')Is this the right formula to register your model with MLFlow backend tracking?
Click on the arrows to vote for the correct answer
A. B.Correct Answer: A.
Option A is CORRECT because it is the correct formula for registering a scikit-learn model with mlflow for tracking.
The <sklearn> must be used as “model flavor”.
Option B is incorrect because the formula actually does the registration of your trained model for being tracked with MLFlow.
References:
Yes, the given code is correct to register a trained ML model with the MLflow backend tracking server.
The code uses the mlflow.sklearn.log_model function to register a trained model with the MLflow tracking server. The function logs the model as an artifact and stores it in the specified artifact path, in this case, "trained_model". Additionally, it registers the model with the specified name, "my_trained_model", on the tracking server so that it can be easily retrieved and shared with others.
The mlflow.sklearn.log_model function is a part of the MLflow Python API, which provides a simple way to track and manage machine learning experiments, including model training, deployment, and versioning.
Overall, the given code is a correct and efficient way to register a trained model with the MLflow backend tracking server.