Skip to content

Update adapter routing for DBX to not require string#19339

Merged
smoorjani merged 3 commits intomlflow:masterfrom
smoorjani:gwt-mlflow-fix-dbx-adapter-route
Dec 16, 2025
Merged

Update adapter routing for DBX to not require string#19339
smoorjani merged 3 commits intomlflow:masterfrom
smoorjani:gwt-mlflow-fix-dbx-adapter-route

Conversation

@smoorjani
Copy link
Collaborator

@smoorjani smoorjani commented Dec 11, 2025

🛠 DevTools 🛠

Open in GitHub Codespaces

Install mlflow from this PR

# mlflow
pip install git+https://github.com/mlflow/mlflow.git@refs/pull/19339/merge
# mlflow-skinny
pip install git+https://github.com/mlflow/mlflow.git@refs/pull/19339/merge#subdirectory=libs/skinny

For Databricks, use the following command:

%sh curl -LsSf https://raw.githubusercontent.com/mlflow/mlflow/HEAD/dev/install-skinny.sh | sh -s pull/19339/merge

Related Issues/PRs

#xxx

What changes are proposed in this pull request?

Update adapter routing for DBX to not require string

How is this PR tested?

  • Existing unit/integration tests
  • New unit/integration tests
  • Manual tests
from unittest import mock

from mlflow.genai.judges.adapters.databricks_managed_judge_adapter import (
    DatabricksManagedJudgeAdapter,
)
from mlflow.genai.judges.adapters.databricks_serving_endpoint_adapter import (
    DatabricksServingEndpointAdapter,
)
from mlflow.genai.judges.adapters.gateway_adapter import GatewayAdapter
from mlflow.genai.judges.adapters.litellm_adapter import LiteLLMAdapter
from mlflow.types.llm import ChatMessage

# Test cases: (model_uri, prompt_type, expected_without_litellm, expected_with_litellm)
TEST_CASES = [
    ("databricks", "string", DatabricksManagedJudgeAdapter, DatabricksManagedJudgeAdapter),
    ("databricks", "list", DatabricksManagedJudgeAdapter, DatabricksManagedJudgeAdapter),
    ("databricks:/my-endpoint", "string", DatabricksServingEndpointAdapter, DatabricksServingEndpointAdapter),
    ("databricks:/my-endpoint", "list", DatabricksServingEndpointAdapter, DatabricksServingEndpointAdapter),
    ("endpoints:/my-endpoint", "string", DatabricksServingEndpointAdapter, DatabricksServingEndpointAdapter),
    ("endpoints:/my-endpoint", "list", DatabricksServingEndpointAdapter, DatabricksServingEndpointAdapter),
    ("openai:/gpt-4", "string", GatewayAdapter, LiteLLMAdapter),
    ("openai:/gpt-4", "list", None, LiteLLMAdapter),
    ("anthropic:/claude-3-5-sonnet-20241022", "string", GatewayAdapter, LiteLLMAdapter),
    ("anthropic:/claude-3-5-sonnet-20241022", "list", None, LiteLLMAdapter),
]


def get_applicable_adapter(model_uri: str, prompt):
    adapters = [
        DatabricksManagedJudgeAdapter,
        DatabricksServingEndpointAdapter,
        LiteLLMAdapter,  # Checked BEFORE GatewayAdapter
        GatewayAdapter,
    ]

    for adapter in adapters:
        if adapter.is_applicable(model_uri, prompt):
            return adapter
    return None


def test_scenario(scenario_name: str, litellm_available: bool):
    print(f"\n{'='*70}")
    print(f"{scenario_name}")
    print(f"{'='*70}\n")

    string_prompt = "test input"
    list_prompt = [ChatMessage(role="user", content="test")]

    passed = 0
    failed = 0

    for model_uri, prompt_type, expected_without, expected_with in TEST_CASES:
        expected_adapter = expected_with if litellm_available else expected_without
        prompt = string_prompt if prompt_type == "string" else list_prompt

        with mock.patch(
            "mlflow.genai.judges.adapters.litellm_adapter._is_litellm_available",
            return_value=litellm_available,
        ):
            actual_adapter = get_applicable_adapter(model_uri, prompt)

        status = "✓" if actual_adapter == expected_adapter else "✗"
        expected_name = expected_adapter.__name__ if expected_adapter else "None"
        actual_name = actual_adapter.__name__ if actual_adapter else "None"

        if actual_adapter == expected_adapter:
            passed += 1
            print(f"{status} {model_uri:40} {prompt_type:6} -> {actual_name}")
        else:
            failed += 1
            print(
                f"{status} {model_uri:40} {prompt_type:6} -> Expected: {expected_name}, Got: {actual_name}"
            )

    print(f"\nResults: {passed} passed, {failed} failed")
    return failed == 0

success1 = test_scenario("Scenario 1: WITHOUT LiteLLM (fallback to GatewayAdapter)", False)
success2 = test_scenario("Scenario 2: WITH LiteLLM (LiteLLMAdapter handles all)", True)

print(f"\n{'='*70}")
if success1 and success2:
    print("✅ All adapter selection tests passed!")
    return 0
else:
    print("⚠️  Some tests failed!")
    return 1
Screenshot 2025-12-15 at 9 17 08 PM

Does this PR require documentation update?

  • No. You can skip the rest of this section.
  • Yes. I've updated:
    • Examples
    • API references
    • Instructions

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.

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

Components

  • area/tracking: Tracking Service, tracking client APIs, autologging
  • area/models: MLmodel format, model serialization/deserialization, flavors
  • area/model-registry: Model Registry service, APIs, and the fluent client calls for Model Registry
  • area/scoring: MLflow Model server, model deployment tools, Spark UDFs
  • area/evaluation: MLflow model evaluation features, evaluation metrics, and evaluation workflows
  • area/gateway: MLflow AI Gateway client APIs, server, and third-party integrations
  • area/prompts: MLflow prompt engineering features, prompt templates, and prompt management
  • area/tracing: MLflow Tracing features, tracing APIs, and LLM tracing functionality
  • area/projects: MLproject format, project running backends
  • area/uiux: Front-end, user experience, plotting, JavaScript, JavaScript dev server
  • area/build: Build and test infrastructure for MLflow
  • area/docs: MLflow documentation pages

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

  • 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/breaking-change - The PR will be mentioned in the "Breaking Changes" 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

Should this PR be included in the next patch release?

Yes should be selected for bug fixes, documentation updates, and other small changes. No should be selected for new features and larger changes. If you're unsure about the release classification of this PR, leave this unchecked to let the maintainers decide.

What is a minor/patch release?
  • Minor release: a release that increments the second part of the version number (e.g., 1.2.0 -> 1.3.0).
    Bug fixes, doc updates and new features usually go into minor releases.
  • Patch release: a release that increments the third part of the version number (e.g., 1.2.0 -> 1.2.1).
    Bug fixes and doc updates usually go into patch releases.
  • Yes (this PR will be cherry-picked and included in the next patch release)
  • No (this PR will be included in the next minor release)

@github-actions github-actions bot added v3.7.1 area/evaluation MLflow Evaluation rn/none List under Small Changes in Changelogs. labels Dec 11, 2025
@github-actions
Copy link
Contributor

github-actions bot commented Dec 11, 2025

Documentation preview for 2efb2ba is available at:

More info
  • Ignore this comment if this PR does not change the documentation.
  • The preview is updated when a new commit is pushed to this PR.
  • This comment was created by this workflow run.
  • The documentation was built by this workflow run.

@smoorjani smoorjani added v3.8.0 and removed v3.7.1 labels Dec 12, 2025
Signed-off-by: Samraj Moorjani <samraj.moorjani@databricks.com>
Signed-off-by: Samraj Moorjani <samraj.moorjani@databricks.com>
@smoorjani smoorjani force-pushed the gwt-mlflow-fix-dbx-adapter-route branch from c542fad to b7fb36d Compare December 16, 2025 05:16
Copy link
Collaborator

@AveshCSingh AveshCSingh left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Left one tiny request, otherwise LGTM

.
Signed-off-by: Samraj Moorjani <samraj.moorjani@databricks.com>
@smoorjani smoorjani enabled auto-merge December 16, 2025 21:43
@smoorjani smoorjani added this pull request to the merge queue Dec 16, 2025
Merged via the queue into mlflow:master with commit 01e8d1d Dec 16, 2025
46 checks passed
@smoorjani smoorjani deleted the gwt-mlflow-fix-dbx-adapter-route branch December 16, 2025 22:24
WeichenXu123 pushed a commit to WeichenXu123/mlflow that referenced this pull request Dec 19, 2025
Signed-off-by: Samraj Moorjani <samraj.moorjani@databricks.com>
WeichenXu123 pushed a commit that referenced this pull request Dec 19, 2025
Signed-off-by: Samraj Moorjani <samraj.moorjani@databricks.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

area/evaluation MLflow Evaluation rn/none List under Small Changes in Changelogs. v3.8.0

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants