How to get url of mlflow logged artifacts?

Question

I am running an ML pipeline, at the end of which I am logging certain information using mlflow. I was mostly going through Databricks' official mlflow tracking tutorial.

import mlflow
import mlflow.sklearn
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_squared_error

with mlflow.start_run():
  n_estimators = 100
  max_depth = 6
  max_features = 3
  # Create and train model
  rf = RandomForestRegressor(n_estimators = n_estimators, max_depth = max_depth, max_features = max_features)
  rf.fit(X_train, y_train)
  # Make predictions
  predictions = rf.predict(X_test)
  
  # Log parameters
  mlflow.log_param("num_trees", n_estimators)
  mlflow.log_param("maxdepth", max_depth)
  mlflow.log_param("max_feat", max_features)
  
  # Log model
  mlflow.sklearn.log_model(rf, "random-forest-model")
  
  # Create metrics
  mse = mean_squared_error(y_test, predictions)
    
  # Log metrics
  mlflow.log_metric("mse", mse)

When I run the above block of code in Databricks notebook, the below status message shows:

(1) MLflow run
Logged 1 run to an experiment in MLflow. Learn more

And I can view the logged information by clicking on "1 run."

However, I would like to automatically retrieve this link. In particular, I need the link to the mlflow uri where the artifacts are stored. This link is in the following format:

https://mycompany-dev.cloud.databricks.com/?o=#mlflow/experiments//runs/

I tried investigating the url and finding the various id codes that are present in it by printing the following information:

print("Tracking URI: ", mlflow.get_tracking_uri())
print("Run id:", run.info.run_id)
print("Experiment:", run.info.experiment_id)

I figured out that in the link above is the experiment_id and is the run_id. But I have no idea what stands for. Also, I believe there should be a built-in functionality to retrieve the link of saved artifacts, instead of manually having to build up the link from sections. However, I haven't found such a funcitonality in the documentation so far.

Edit: Now I discovered that is the Databricks workplace id. But it is still a question how I can access it programatically.

How to get url of mlflow logged artifacts?

0 Answers