[Custom Metrics] Enabled support for logging of numerical metrics

[MarkYHZhang]

Signed-off-by: Mark Zhang mark.zhang@databricks.com

What changes are proposed in this pull request?

Added support for users to pass in custom metric functions into mlflow.evaluate. This PR is a part of more PRs to come for custom metrics support. Specifically, this PR focuses on enabling logging of numerical metrics generated by user's custom metric functions into MLflow.

How is this patch tested?

Wrote unit tests located in mlflow/tests/models/test_default_evaluator.py

Does this PR change the documentation?

Updates to the documentation will be done once the entire feature of custom metrics is complete.

• No. You can skip the rest of this section.
• Yes. Make sure the changed pages / sections render correctly by following the steps below.

1. Check the status of the ci/circleci: build_doc check. If it's successful, proceed to the
   next step, otherwise fix it.
2. Click Details on the right to open the job page of CircleCI.
3. Click the Artifacts tab.
4. Click docs/build/html/index.html.
5. Find the changed pages / sections and make sure they render correctly.

Release Notes

Is this a user-facing change?

• No. You can skip the rest of this section.
• Yes. Give a description of this change to be included in the release notes for MLflow users.

Enables MLflow tracking of custom metrics produced by user-provided metric functions.

(Details in 1-2 sentences. You can just refer to another PR with a description if this PR is part of a larger change.)

What component(s), interfaces, languages, and integrations does this PR affect?

Components

• area/artifacts: Artifact stores and artifact logging
• area/build: Build and test infrastructure for MLflow
• area/docs: MLflow documentation pages
• area/examples: Example code
• area/model-registry: Model Registry service, APIs, and the fluent client calls for Model Registry
• area/models: MLmodel format, model serialization/deserialization, flavors
• area/projects: MLproject format, project running backends
• area/scoring: MLflow Model server, model deployment tools, Spark UDFs
• area/server-infra: MLflow Tracking server backend
• area/tracking: Tracking Service, tracking client APIs, autologging

Interface

• area/uiux: Front-end, user experience, plotting, JavaScript, JavaScript dev server
• area/docker: Docker use across MLflow's components, such as MLflow Projects and MLflow Models
• area/sqlalchemy: Use of SQLAlchemy in the Tracking Service or Model Registry
• area/windows: Windows support

Language

• language/r: R APIs and clients
• language/java: Java APIs and clients
• language/new: Proposals for new client languages

Integrations

• integrations/azure: Azure and Azure ML integrations
• integrations/sagemaker: SageMaker integrations
• integrations/databricks: Databricks integrations

How should the PR be classified in the release notes? Choose one:

• rn/breaking-change - The PR will be mentioned in the "Breaking Changes" section
• rn/none - No description will be included. The PR will be mentioned only by the PR number in the "Small Bugfixes and Documentation Updates" section
• rn/feature - A new user-facing feature worth mentioning in the release notes
• rn/bug-fix - A user-facing bug fix worth mentioning in the release notes
• rn/documentation - A user-facing documentation change worth mentioning in the release notes

[MarkYHZhang]

Hey @apurvakoti @dbczumar @jinzhang21 @WeichenXu123, can you all take a look? It seems like I don't have permission to add reviewers. So I'm tagging you all here. Thanks!

[jinzhang21]

Added some initial comment on the API. I'll review the rest later today.

[apurva-koti]

Some small comments, will do a full review again

[apurva-koti]

Looks good! Let's use [] as the default argument for custom_metrics rather than None.

[MarkYHZhang]

Looks good! Let's use [] as the default argument for custom_metrics rather than None.

Thanks! I think we should avoid using mutable default argument in python due to issues described here. Although, I could initialize it to a list inside the constructor, but considering other default arguments like

mlflow/mlflow/models/evaluation/base.py

Line 655 in 2c0d13d

feature_names: list = None,

are not initialized, I think it might be good idea just to leave it as is.

[apurva-koti]

Ah good catch @MarkYHZhang ! Let's keep it None then.

[apurva-koti]

LGTM once all comments addressed!

[dbczumar]

@MarkYHZhang This looks great! Left a few small comments - should be ready to merge soon!

[dbczumar]

@MarkYHZhang Do we plan to add an example to the examples section or extend one of the existing examples? Can we do that as part of this PR?

[dbczumar]

LGTM when small docs nits have been addressed and example has been included. Thanks @MarkYHZhang ! Awesome stuff!